A Study Nasa Cr-1532 Tech Library Kafb, Nm

A STUDY NASA CR-1532 TECH LIBRARY KAFB, NM

A STUDY OF PROTON-INDUCED EFFECTS

ON REFLECTIVE SURFACES OF SPACE MIFtR0R.S

By Roger B. Gillette and Bruce A. Kenyon

Distribution of this report is provided in the interestof informationexchange. Responsibility for thecontents resides in the author or organization that prepared it.

Prepared under Contract No. NAS 1-7627 by THE BOEING COMPANY Seattle, Wash.

for Langley Research Center

NATIONAL AERONAUTICS AND SPACE ADMINISTRATION

For sole by the Clearinghouse for Federal Scientific and Technical Information Springfield, Virginia 22151 - CFSTI price $3.00

- TABLE OFCONTENTS

2 Page

1.0 SUMMARY 1

2.0 INTRODUCTION 3

TEST SPECIMENS 3.0 TEST 6

4.0 APPARATUS AND PROCEDURES 7 4.1 Proton Radiation FacilityRadiation Proton 4.1 7 4.2 In-Situ OpticalIn-Situ 4.2 Measurement Facility 9 4.2 .1 Design Requirements Design 4.2.1 9 4.2 .2 Ultraviolet ReflectometerUltraviolet 4.2.2 9 4.2 .3 Scattered-Light4.2.3 Measurement Apparatus12 4.3 Visible and Infrared ReflectometersInfrared andVisible 4.3 13 4.4 Interferometers 14

5.0 RESULTS AND DISCUSSION 15 5.1 MgF /Aluminum CoatedMirrors 15 2 5.1.1 DependenceIntegrated-Flux 15 5.1.2 Flux Dependence 16 5.1.3 Post-IrradiationReflectance Changes 17 5.1 .4 Surface Finish Studies Finish Surface 5.1.4 18 5.1.5 Scattered-Light Data 19 5.2Coated LiF/Aluminum 21 Mirrors 5.2.1 Integrated-Flux Dependence 21 DependenceIntegrated-Flux 5.2.1 5.2.2 Flux Dependence 21 Dependence Flux 5.2.2 5.2 .3 Post-Irradiation5.2.3 Reflectance Changes 22 D iscus sion of Results 23 Results of 5.3 Discussion 5.3 .1 Contamination5.3.1 DetectionExperiments 23 5.3 .2 Residual 5.3.2 Gas Analyses 26 5.3.3 ContaminantDeposition5.3.3 Film Rate 27 5.3.4 Cleaning Experiments Cleaning 5.3.4 28 5.3.5 Contaminant-Film-InducedPrediction of 31 Reflectance Changes 5.3.6 Degradation 5.3.6 34 in Space

6.0 CONCLUSIONS AND RECOMMENDATIONS 35

7.0 APPENDIXES AppendixA-Mirror Sample Procurement Specifications 37 AppendixB-Miscellaneous Data On Mirror Substrates & Coatings 40 Appendix C-Optical Constants of Aluminum and MgF 43 2 AppendixD-Optical Constants Contaminant for Film46

8.0 RE3EFXNCES 47 -

~ ~~~~~ d iii Q A STUDY OF PROTON-INDUCED EFFECTS ON REFLECTIVE SURFACES OF SPACEMIRRORS

By Roger B. Gillette andBruce A. Kenyon TheBoeing Company, Seattle, Washington

1.0 SUMMARY

The results of a research program to study the effects oflow energy protons on reflective surfaces for space optical systems, are presented in this report. Theprimary objective of the program was todetermine the effects on telescope mirror reflective surfaces, of the proton radia- tionenvironment at synchronousEarth orbit. Secondary objectives of the study were to obtain information on surface finish characteristics and reflectance characteristics of three promising mirror substrates.

Mirror test: specimensevaluated in thestudy included polished substrates offused silica, Cer-Vit*, andKanigen**-nickel-plated beryllium; coatedwith aluminumand overcoated with MgF Also included were-speci- 2' mens ofpolished fused silica coatedwith aluminumand LiF. The experimental program studied the dependence of specular reflectance and scatter- ingproperties on: (1) proton integrated flux; (2) protonflux; (3) post- irradiationexposure to air; and (4) coating and substratetype. The effects of integrated fluxes up to 10l6 protons-cm-2 sec-' of 10 keV energy were evaluated.

Results of experiments showed that the primary mechanism of damage was proton-induced deposition of a contaminant film onto the mirror surfaces.Cleaning experiments and theoretical predictions of reflectance changeson MgF /A1-coated mirrors, showed that the contaminant film 2 accountedfor essentially all of theobserved damage. Similarlycleaning experiments on LiF/Al-coated mirrors indicated that the majority of the observed damage was a contaminantfilm effect. In relating experimental results tothe synchronous-orbit space environment, it was concludedthat negligible degradation will occur over a two-yearperiod if no contaminant film deposition occurs.

An oxidation cleaning technique was developed which could be used for removing the film and restoring the reflectance to nearly the pre- irradiationvalue. Results suggested that this cleaning technique could be utilized for salvaging space mirrors which may become contaminated in either Earth tests or space use.

~~ * Cer-Vit is thetrade name for a lowcoefficient of expansion I material produced by Owens-Illinois Company. ** Kanigen is theGeneral American TransportationCorporation trade name for a non-electrical technique of platingnickel-phosphorous.

4 Lightscattering measurements at wavelengthsof 253.7 nm and 500 nm revealed no significantradiation-induced changes in scattering from eithertype of surface. Surface fin$sh studies also showed thatthe fused silica and Cer-Vit substrates were polishedsmoother than beryllium substrates, and the deposition of reflective coatings had a negligible effect on microroughness. 2.0 INTRODUCTION

The primary objective of this study was to determine the effects of the proton radiation environment-at synchronous Earth orbit, on telescope mirror reflective surfaces.Secondary objectives of the study were to obtain information. concerning. the surface finish and reflectance characteristics of.threepromising mirror substrates. It is intendedthat the results will be of assistance in choosing the mirror substrate and coating combination which will provide the optimum combination of high reflectance andlow degradation. inthe space environment. The radiakion environmentand mirror test specimens were selected for compatibility withplans (ref. 1) toplace a largeastronomical telescope mirror at synchronousaltitude. However, thestudy results should be applicable to many other space optical 'systems (ref. 2).

Mirror test specimensevaluated in the study included polished substrates offused silica, Cer-Vit, andKanigen-nickel-plated beryllium; all were coatedwith aluminumand overcoatedwith MgF . Also included were specimens of polished fused siliea coaced with azuminum and overcoated withLiF. These vacuum depositedcoatings were appliedwith techniques which-produce maximum reflectance in the vacuum-ultraviolet wavelength region.Uncoated, polished specimens of each substratetype were also prepared for surface fipish experiments.

The, exper.imenta1. program studied- the dependence of specular refleetanceand,scattered light properties on: (1) .protonintegrated flux; (2) protonflux;- (3) post-irradiationexposure to air; and (4) substrate and-coatingtype. The effectof the vacuum depositedcoatings on sub- stf%te,micro-roughness was alsoevaluated. T;roton integrated flux s to 10 protons-cur -2 andfluxes between 2 x 10 and 1OI2 protons-cm- 2 sec-1 were used,Flux dependence was evaluated to establish the validity of testing at exposure rates far higher than those encountered in space. Post-irradiation reflectance measurements(before and after exposure to air) were performed- to determine- whether specimens of this type need to be held,in vacuum in the time period between irradiation and reflectance measurements.Reflectance measurements were performedover the wave- lengthrange.from 90 to 50,000 nm. Theregion from 90 to 250 nm was covered with an ultraviolet reflectometer which provided capability for performing both reflectance and scattered light measurements in the irradiation chamber. Reflectance at longerwavelengths was-measured in moreconventional, commercial reflectometers.

At the beginning of the study an analysis. was .made of the most recent charged-particleenvironment .data at synchronousaltitude. Solar electromagnetic radiation was not considered because planned configurations for telescopemirrors were shieldedfrom the sun. The trappedproton environmentsuggested for this study is presentedin Figure 1. Boththe integralflux (f) and thedifferential flux (df/dE)*spectra of these trappedprotons are shown. These spectra are basedprimarily upon the - lowenergy results reported-by Frank (refs. 3 and 4) and are consistent withthe results of Katz, et al, (ref. 5). Theaverage omnidirectional * (E) representsproton energy I- 3 integral flux of protons is expected to be 3.5 x 108 protons-cm'2 sec-' or 2.8 x lo7 protons-cmr2 sec ster-I. A surfacewith 27r geometry will encounter 5.5 x 1015 protons cm-2 year-l with roughly 70 percent having energies less than 30 keV.

The trappedelectron environment-at synchronous altitude, based upon theresults presented.by Frank (ref. 3) and Vette (ref. 6), yields a time averagedelectron flux of about 1 x electrons-cm'2year-1 withenergies greater than 10.keV on a 27r surface. The highenergy component is adequately re resented.by an exponential spectrum with an integral flux of. 1.5 x lOlg electron-cm-2year-l having a mean energy of 215 keV (ref. 6).

The surface ionization dose. for. the combined proton and electron environment was estimatedto be about 2 x 108 Joules-kg-l-yr-l, with about 5 percentbeing due to the electrons. Displacement damage inthe overcoating would bepredominantly proton induced. Based onthese considera- tions; this study was directedtowards determining the effects of low energy protons.

In selecting a proton energy for this program, consideration was given to the.ionization dose and displacement damage expectedfrom the aboveenvironment,. The differential surface dose rate was found to peak between-15and-40 keV, while the displacement yield was largely due to theprotons below 10 keV. Thus,an energy of 10 keV was chosenfor the tests. It was anticipatedthat this energy would best simulate the competition-of displacement and ionization damagemechanisms occurring at synchronousaltitude.

Results of radiation effects-studies on specularly reflecting surfaceshave been reported bynumerous authors (refs. 7-13). Inthese references,primary emphasis was placed on-specular surfaces to be used in solar concentrators and forspacecraft thermal control. Several flight experiments are presently in progress in which mirror specimens are beingevaluated in space. These include the Air Force OV1-10 vehicle, and NASA OSO-3 and..ATS-3 vehicles.Mirrors having high ultraviolet reflectance were not included in these or similar earlier experiments.

At the outset of this programonly a limited amount of radiation effects studies had been done on vacuum-ultraviolet reflecting optical surfaces.In studies by Canfield, et al. (ref. 14), MgF2-overcoated aluminum films were irradiated with 1-MeV electrons, 5-MeV protons,and ultravioletradiation. Lt was reportedthat negligible changes in reflectance occurred at 121.6 nm as a result of exposures to the above typesof radiation. It was apparentfrom available literature that inadequate data existed for predicting degradation of mirrors in a synchronous-orbitproton environment. Additional information was needed on: (1) theeffects of low energy protons; (2) theeffects ofhigh integratedfluxes; (3) theeffects of irradiating at a rate higherthan thatexperienced in space; and. (4) the effects of air exposureprior to performingpost-irradiation reflectance measurements. It was of^ primary interest to relate the above effects to specular reflectance changes Over the wavelength region in which space mirrors will be used.

4 Special acknowledgement is givento Dr. Georg Hass of the U. S. Army Electronics Command, NightVision Laboratory, and Mr. William R. Hunter of the Naval Research Laboratory for preparation of certain test specimens, numerous helpfulsuggestions, and supplying optical properties data. Acknowledgement is alsoextended to Mr. ReynoldWilbert for experimental equipmentdesign and preparation, Mr. SheridanCannaday for assistance in conductingradiation experiments, Dr. William Dohertyfor space environment andradiation effects analyses, Dr. Sarojini Das forperforming interferometricmeasurements, Mr. William Woods for electronic design and for analysis andprogramming assistance in the theoretical reflectance calculations, Mr. Melvin Horman for reflectometer aberrations calculations, and Mr. EdwardHoffman of NASA/Langley Research Center for program technicalguidance.

5 3.0 TEST SPECIMENS

The mirror test specimenschosen for study in this program were typicalof those used in ultraviolet reflecting optical systems. The basic requirements for those mirrors are that they maintain high reflectance down to at least 120 nm wavelength,and that they be sufficiently polished to minimize scattered light.

Threedifferent types of substrates were evaluated.These were: Corningfused silica #7940, Cer-Vit (Owens-Illinois premium grademirror blank material C-lOl), andKanigen-nickel-plated beryllium (Berylco grade HP-40). Two differentcoating systems were evaluated; MgF overcoated .aluminum,and LiFovercoated aluminum.The MgF2 overcoatZng was appliedto all threetypes of substrates, whereas, the LiF overcoating was appliedonly to a batchof fused silica substrates.LiF overcoated mirrors were included because of their ability to maintain high reflectance down to a wavelengthof about 102 nm (MgF beginsto absorb 2 at wavelengths shorter than about 130 nm).

The detailed specification used for procurement of mirror test specimens is given in Appendix A, and a briefdescription follows: The mirrors were nominally5.08 x 5.08 cm squareand 1.90 cm thick. The front surface was to be figured to 1/10 for the mercury green line (546.1 nm). Surfacepolish was to beadequate for minimizing scattered light in the vacuum-ultraviolet region with the exception of the Kanigen-nickel-platedberyllium, which was polishedon a best-effort basis.In the case ofthe MgF2 overcoatedsurfaces, reflectance was to be maximized at 121.6 nm in accordance with coating thicknesses and proceduresgiven by Canfield, et al. (ref. 14). In the case ofLiF overcoated surfaces, reflectance was to be maximized at 102.6 nm using proceduresgiven by Cox, et al. (ref. 15). Mirrorsubstrates were preparedand polished by TinsleyLaboratories. MgF /aluminum coatings 2 were applied by OpticalCoating Laboratory of Santa Rosa, California andLiF/aluminum coatings were applied by Dr. Georg Hass and associates at the U. S. Army Electronics Command, NightVision Laboratory, Ft . Belvo ir , Virginia. Detailedinformation regarding chemical and physical properties of the specific batches of substrate materials is given in Appendix B. Data obtainedfrom the vacuum coating laboratories on cleaning andcoat- ing procedures is also given in Appendix B.

Vacuum deposited film thickness measurements were performed on micro- scopeslides which were coatedalong with the MgF overcoatedspecimens. 2 The slides were prepared with an uncoated strip to provide a step for interferometermeasurements. The entire surface was thenovercoated with about 15 nm ofaluminum toobtain uniform optical properties. Results of thesemeasurements indicated an average total film thickness of about 115 nm on the beryllium and Cer-Vit mirrors, and an average of about128 nm onthe fused silica mirrors.These values are in agreement with the film thicknesses specified in the procurement specification (601120 nm of Al and 25 nm of MgFZ). W 6 4.0 APPARATUS AND PROCEDURES

Includedrin this section are detailed descriptions of the proton radiation facility, the ultraviolet reflectometer; the scattered-light measurement apparatus, the visible and .infrared reflectometers, and the interferometersutilized in this study. In accordance with the program requirement that reflectance be measurable without exposing specimens ta air after irradiation, the first three.above items were combined into a single.in-situ facAlity. The followingdiscussion inc1udes.a general description of the overall radiation test facility, and detailed descriptions of the ultraviolet reflectometer and scattered light measurement apparatus.Discussions on experhental and operational procedures, where applicable, are integrated.with descriptions of the apparatus.

4.1 ProtonRadiation Facility

A schematic of the overall-radiation test faeility is shown in Figure 2.The major items of' the facility are the proton source (accelerator) , the irradLation test chamber,and the ultraviolet monochromator.Protons entered the test chamber thr0ugh.z' 5-cm diameterhole in the collimating.mirror, and impingedon the test mirror at the opposite endof the chamber. To performreflectance and scatteredlight measurements,light from the monochromator* enteredthe optical system through an aperture behind and to the side of thespecimen.

An ORTEC-kkrf-excited ion.source was used an the accelerator for generatingprotons from hydrogen^ gas. Eleetrostatic-lenses were usedfor accelerating. and. focusing the proton beam.. Separationof protons from mass 2 and-3 'ions was achieved.with an electromagnet.

The. test specimen .was mounted inr,a. fixture which exposed a 3.81-cm (1.500-in)diameter area tothe proton beam (Figure3). The exposed area was defined by a thin; stainless- steel aperture located 0.025-cm in front of the specimen, Thisarrangement provided a sharpdefinition between the irradiated and.non-Lrradiated areas to aid interpretation of results in subsequent interferometric examination,

Proton flux was m'easured with an array of three Faraday cups which traversedthrough the beam just ahead of the test mirror. The Faraday cups,having apertures of O.476-cmy scannedacross the.beam at the top, bottom,and-center of the 3.81-cm diameteraperture. Flux readings taken at variouspositions within the irradiated area were averaged.Uniformity of the proton flux over the irradiated area was maintained within --+ 5 percent for most- tests. During long term exposures, readings were taken about, every half hour and integrated flux values were calculated from these data.

* McPherson InstrumentCorporation Model225 ** Oak Ridge Technical Enterprises Company 7 Protonenergy was established by theaccelerating voltage in the m ionsource; in this case 10 kV. Beam energy measurements were not performedbecause earlier experiments (ref. 13) had shown thatmeasuring the, accelerating voltage is adequate.

The procedurefor irradiating specimens was as follows: A mirror specimen* was installed in the masking fixture, placed in the chamberand opticallyaligned. Thechamber was evacuatedand scattered light measurements were made in accordance with procedures discussed in a later section. The pre-irradiation ultraviolet reflectance.measurement was then made. During the optical measurements the test chamber was open to the monochromatorand isolated from the accelerator by a gate valve. Typical chamber pressures during reflectance measurements were in the order of 7 x torr.Following completion ofreflectance measurements, the test chamber cold trap was filled with nitrogen and the chamber was closed to the monochromatorand.opened to the accelerator beam tube. Duringtuning of the accelerator the Faraday cup array and attached shieldprotected the test mirror. When a satisfactoryproton flux and uniformity were obtained, the Farday cup was moved to a park position and irradiation was begun.Typical chamber pressuresduring irradiation were 2-5 x 10-7 torr when operating with a flux of 1011 protons-cm’2. sec-l, and 7-9 x 10-7 torr for a flux of 10l2 protons-cm-2 sec-1. It shouldbe noted tha,t- residual gas analyses indicated hydrogen to be the pre- dominantgas in the chamber. Sinceion gauges are normallycalibrated for air, thesepressures are thereforesubject to some error.

At selected levels ofintegrated flux, irradiat,ion was interrupted by closing the beam-tube gate valve, and reflectance measurements were performed. Many of thelow-flux exposures were completedover a period of several days. In these cases, the test was shut ‘down at nightafter completing a reflectance measurement.Preliminary experiments had shown that negligible changes in reflectance occurred during such overnight shut-down periods.

As soon as possiblefollowing completion of irradiation, the reflectance and scattered-lightdistribution were measured.The chamber was thenbackfilPed with ambient air. Some specimens were allowedto stand in air for periods of time after irradiation to determine whether any post-irradiation reflectance and scattered-light changes would occur. Thechamber was re-evacuated for the optical measurements,

It was assumed at the outset of the program that contamination of test specimensmight be a problem.Therefore, care was taken in selection of materials to beused in the radiation facility to minimizethe use oforganic compounds. An ideal facility would havebeen fabricated with non-organiq bakeable materials and all pumps wouldhave beenion pumps. However, thecost of such a systerh seemed prohibitive.

* Prior to this step, pre-irradiation reflectance measurements in the near-ult~raviolet, visible, and infraredwavelength regions hadbeen completed. v 8 - No suchmonochromator was commercially available, and extensive modifica- tionof the existing1proton accelerator would 5ave beenrequired. In conkideration of such factors, compromises were made on the use of organic materials.

The-system assembled for this study utilized stainless steel vacuum chambersand seals of both .Viton-A.O-rings and rnetal. gaskets. Apiezon-L grease was usedfor lubrication of. O-rings, The accelerator beam tube was evacuated with an ion-pump, the ‘test chamber with a Turbomolecular* pump, and-the monochromator with a diffusion pump. The diffusion pump was operated with DC-704 silicone oil and was trapped with a liquid nitrogen.chevronbaffle. No coldtrap.was employedbetween the Turbo- molecular pump and the test.chamber, although a cold-finger type trap was used- in thechamber.

4.2 In-SituOpt-icd Meanuement Facility

The in-situ optical meazarene~facility cons$.stedof. an ultraviolet reflectometer-and apparahs for measur.ing changes in seattffed light from mirror. specimen-s.The following .seetion includes discussions of general designrequirements the ultraviolet reflectometer, and thescattered light apparatus, in respective order.

,4.2.1 Design- RequireWs.- Proton-induced damage on telescope mtrrors. may result in efther of two types .of.optkcal. degradation: (1) an increase in the .amount of scattered light; or (2) a change in reflectance produced-by increased abssrptton or by shifted interference effects. An increase in scattered light will reducetelescope image contrast; a decrease in reflectance makes the detection of faint objects more difficult. Measurementsof both scattering and reflectance properties are necessary to predict ehe type andamount of degradation

I expectedof a spacemirror.

It was assumedfrom rhe outset that reflectance measurements.should be made in-situ, i.e.; immediately after irradiation without exposure to air. Impdsitionof this design requirement was basedon the numerous oxygen-induced recovery effects-which have been noted in radiation effects studiesof spacecraft thermal control coatings. The initial equipment designdictated that irradiated specimens be exposed to air beforeper- formingscattered light measurements, however, final equipment modifications allowed measurements to be made in-situ.

4.2.2 UltravioletReflectometer.- A schematicshowing the.reflectometer and scattered-light measurement apparatus,isgiven in Figure 4. The general arrangement was patterned after the Ebert-Fastiemonochromator, withthe specimen in the grating position. The beam from a 1-meter. vacuum-ultraviolet,monochromator (McPhersonModel 225) expands until it - strikesthe spherical collimating mirror. At thispoint, the spherical

. * Trade name for a vacuum pump manufactured by Welsh Scientific Company. 9 mirror. rediracts .the light to the specimen.Light reflected from the specimen.illuminatesthe opposite side of the collimating mirror and is c thenfocused on.the exit aperture.In the absolute reflectance measurement mode, aninternal* photomultiplier (PM) tubecan be positioned to intercept the light either before it reachesthe specimen or after it is reflectedfrom the specimen ("incident" and "reflected"positions, respectively). With careful management ofthe. system aperture stop to accountfor the optical non-equivalence of the two positions, the ratio of outputs then yields the absolute reflectance of the specimen.

The reflectometer was designedto cover the wavelength range from 90 to 250 nm. To obtainsufficient power foroperation ofphotomultiplier (PM3 tubes in this range, a gratingblazed at 15.0 nm was selectedand mirrors were overcoated with LiF to get.high.ref1ectance down to about 100 nm; A Hinteregger-typelight source for the monochromator was operated windowless with a d.c. hydrogen discharge.

Several small modifications were made to the monochromator for its use withthe'reflectometer. These included replacement of the exit slit assemblywith a circular aperture 0.762 mm (0.0300inch) in diameter, installation of an aperture stop about halfway between the grating and exitaperture-,..and installation of a movable filter betweenthe entrance slit and thegrating (Figure 4). The exit aperture, combined withan entrance slit setting of 0,762 m, defined a passbandof about 1.4 nm**. The aperture s.t-op in the monochromatorhad two openings which divided the outputinto side. by side. beams (Figure 4). One openinglimited the size of the collimated beam at the test specimen location to 3.528 cm diameter, and the,other opening defined a beam intercepted totally by the diagonal mirrorand reference PM tube.Light-incident on the test mirror was comp8:etely containedwithin the 3.81-cm diameterradiation mask.The active aperture used on the internal PM tube was 3.56 cm, large enough to accept all of theincident and reflected beam. Theseaperturing techniquescontrolled stray light within the reflectometer, thus per- mitting quite sensitive measurements of scattering distributions.

Themovable filter was installed in the monochromator to correct a problem tha.t* occurred- after several months of operation. The problem involved fluorescence on the spherical collimating mirror which introduced an erroneously high output from the internal PPI tube when it was nearest themirror in the "incident" position. It was determinedthat the fluorescence (resulting in a reduced reflectance in a band at 195 nm) was causedby second-order (i.e., 98 nm) light from thegrating. Theproblem was eliminated by insertion of a sapphire filter in the light beam when measuring.at.wavelengths longerthan 170 nm. The defocusingeffect of the filter on the monochromator is negligible for passbands of the width used in thig program (less-thanone percent increase in passband).This filter also eliminated second-order rad-iation in light from the monochromator at wavelengthsbelow 290 nm.

* "Internal" mans that the tube was. located inside of. the vacuum chamber - ** Measured width at half power. Theoreticalwidth at half power is 1.26 nm; full width is 2.52 nm.

10 For ultraviolet- refle&om&ry, the7ang.k of. incidence on the specimen was 6,056deg. at arc frcm normal. The,monqchromator exit aperture, as noted earlier; was.0.762.mm .in d2m.e. .Its subtense of 3.7 mfnutesof arc at the spherical mirror was the .nomid deeollimation of the incident beam. The.exit aperture for the refleeted beam subtended 6.2 minutes of arc. Adding0.6 minute for astigrpatism, the instrument profile for scattering was 10.5 arc minntes,The.eollima.ting mirror was an f/4.3 spherical reflector with a radius-of .1L&61 cm (55.,75 in.). The effeetivespeed of the test apparatus was €./18..6, as eompared with.f/17.7 for-the largest circular area that can. be iLllumim%z& by thE monochromator'.(The mono- chromatorused a 56-, by 106+mm, 1-meC.er.focus'grating-witha constant 15 0 between incident and :diffracted beams.)

Each.detector assembly-comprised a sedium salicylatefluorescent screenfollowed by a 13-stagepho?zomuftiplier; The internal PM tube was a type EMI 9514 S, and the reference and scattering PM tubes,type 9635 B. An electronic feedback .control circuit was developed for operating the PM tubesduring both reflectance and scattered light measurements. A schematicof the control circuit is shown in Figure 5. All three PM tubes including the reference PM; the internal PM and the scattering PM were suppliedfrom a.single.high voltage source. Voltage variation of the supply was controlled by feedback from the reference PM to hold the PM currentoutput at a constantvalue. In this way, temporalvariations causedby a varying light source output were compensated.The operating point,about which this control was.exercised, was set byproportioning the voltage output from the reference PM tube electrometer (electrometer A).

To make an absolute reflectance measurement the monochromator was set at a desired .wavelength and the internal PM was moved to intercept 1ight.usually incident.on the specimen.The feedback current control was used to vary the power supply voltage to produce a convenient full-scale reading .on electrometer-B which indicated internal PM current (typically 1.00 u A. The internal PM was thenturped to receive the reflected light from thespecimen, and its output current was read as a fraction of 'the full scale readingon electrometer-B. This fraction directly gives the specimen reflectance.

Theuse.of photomultipliers with slightly different gain vs voltage characteristics resulted in a second order variation in the compensation appliedto the internal PM. Temporal variationsin the Lamp output produced..voltage. variations which kept the reference PM tube output currentconstant. Slope differences between the characteristic curves for the reference and internal PM tubes caused minor variations in the currentoutput from the.interna1 PM tube. To guardagainst serious error fromthis cause, each measurement sequence included as its final step a repeatof the initial reading. A major shift'ln the full scale readLng for incident light was interpreted as a light source variation beyond the scope of compensation, requiring a repeat.measurement.

The generalperformance of the reflectometer was good.Reflectance - valuescould be reproduced to within 0.005 reflectance units (for example, 0.800 0.005 or 0.100 2 0.005). The absoluteaccuracy of the reflectance data was not determined.

11 4.2.3 Scattered-Llght.&asurement-Appatatus.- Scattered-light e measurements were performed at wavelengthsof. 253.7and 500 nm. Light sources for scattering measurements were- attached to the monochromator at the, position normally occupied by the Hinteregger- lamp. This arrangement permitted scattering measurements to be madein-vacuum immediately afterreflectance measurements.For visible-light scattering (500 nm), the monochromator was set.at.ze,ro wavelength and the grating used as a mirror. An external,.tungsten-filament lamp,supplied from a constant voltagesource, illuminated khe monochromator entrance slit throughan interference filter. The filter has a 21 nm (half power) pass band centeredat-500 mn. Changeoverfrom reflectance to scattering mode was aecomplished.withoutexposing the specimen to air with the aid of a flap valve in the monochromator. Thisflap valve isolated the entrance slit bodyfrom the remainder. of the equipment.

For scattering measurements at.253.7 m, a lowpressure mercury-vapor dischargetube was mounted insidethe entrance slit body.This lamp is similar in characteristics to the commercial Pen-Ray lamp,but is speciallydesigned to work in a vacuum environment.The monochromator was set at 253.7 nm to isolatethat. radiation. Although the monochromator was operated with a 2.6 nm pass band, the only radiation present in significant amount was the 253.7 nm line.

In the scattered-light measurement mode, the internal PM tube was placed in a park position so,that the beam from the specimencould be collected on the opposite side of the collrimating mirror and focused therebyon the exit aperture. Light that fell within the exit aperture couldthen pass through' and illuminate the scattering PM tube. To measure the scattered-light distribution the specimen was rotated through a small angle (4- 2 deg.of arc). This rotation caused the image to move acrossthe exit-aperture, thus bringing into focus light scattered through twice the angle of- mirror displacement.

For scattering measuremenes, the seattering-signal PM tube was connected to the feedback control circuit in the place of the reference PM tube(Figure 5). The reference PM and the internal PM were both disconnected. Thesystem then functioned as a logarithmicphotometer with the high voltage.analog. output indicating the quantity of-light receivedrbythe scattering PM'tube. A decreasein light input was indicated by the increase in voltage required to keep the PM currentconstant.

To calibrate the photometer,the specimen was rotated to the specularpeak and the input resistance of the electrometer was decreased by decades,This increased the control current bydecades" causing the systemvoltage to increase.. Over thelinear range of the PM tubethe effect is equivalent. to a reduction of the input light by decades while keepLng thecontrol current constant. To extendthe calibration beyond this range, a neutral density filter was inserted in the light path and the process was repeated.

36 The reflectometer provides an output voltage analog Of the current input.

# 12 . To record the.scattering.distribution, thehighyvoltage analog. signal was fedto the y-axis .of an x-y plotter. The x-axis was driven from adpotentiometer circuit directly connected to the cam whichproduced specimen rotation.

Early in the program it was plannedto perform a portion of the reflectance measurements with the scattered-light measurement setup. This arrangement has the basic capability ofmeasuring relative reflectance (bycomparing to readingstaken on tIie internal PM) with strong dis- criminationagainst scattered light*. It was desiredto obtain such data to ascertain whether reflectance changes measured-with the internal PM were influenced by increasedscattering. Although initial experiments included such anti-scattering reflectance measurements, the precision of the data was quitepoor, The technique was abandoned when analysis indicated that scattering from irradiated mirrors created negligible error in reflectances measured.with the internal PM.

4.3 Visible and InfraredReflectometers

Reflectance measurements in the wavelength region from220 to 2500 nm were made with a Cary Instruments Company Model 14 Spectro- photometerusing a specularreflectance attachment (Model 1413). These measurements were made in air. The data obtained with this instrument was absolute specuxar- reflectance and thus could be correlated with data takenwith the .vacuum-ultraviolet reflectometer. In special experiments conducted to measure this correlation, data from the two instruments overlappedin. the wavelength region from 220 nm toabout 400 nm. The agreement was.within 2 percent.

The procedurefor measuring .reflectance with the Cary-14 reflectometer was as follows: All mirrorspecimens including a controlspecimen, were measured inone batch before irradiation. Following irradiation, each specimen was remeasuredalong with the concrol.mirror. The control mirror was included in the reflectance measurements to obtain correctionsfor small systematicerrors. This permitted a more accurate determination of reflectance changes after irradiation.

Infrared reflectance measurements . from 1000 to 50,000 nm wavelength were made with a Beckman XR-12 Spectrophotometer using a specular reflectanceattachment, Data from. thisinstrument is relative, thus, a control specimen had to be run with the two batches .of specimens (before and after irradiation).

* With thespecimen set to image the refleetomezer entrance aperture - ontothe exit aperture, the light reaching the scattering PM includes only a small. amount of scattered light (i.e., that within a cone. of - 9 arc minutes about the angle of regular. reflection). 13 4.4 Interferometers w

Interferometric analyses were made to determine the overall flatness of the,polished mirror surfaces, the thicknesses of vacuum deposited reflective films, the microfinish of thepolished surfaces, and the differencesbetween irradiated and non-irradiated areas. Three differenttypes of interf~erometerswere used.The firsttype, a Fizeau interferometer*, was used for measuring overall flatness on the polished surfaces.Thfs.instrument uses the double beam Fizeaufringes, formed withmercury 54671 nm light, to measure variations in path difference betweenan' NBS calibrated.ref erence flat and the test surface. The secondtype, a Zeiss InterferenceMicroscope, was usedto examine the surfacefinish of themirrors. This instrument- is, inprinciple, a double-beam Michelson interferometer using either white light or mono- chromaticlight (Thallium 535.0 nm). The thirdtype, a Hilger-Watts thin film interferometer, was used~forfilm thicknessmeasurements and examinationof the irradiated mirrors. This instrument utilizes fringes ofequal chromatic order.

Thegeneral procedure used in interferometer measurements was to overcoat the rest mirrors with a thin layer ofaluminum to provide both uniformoptical properties across the surface, and sharperfringes. A minimum of three similar measurements were taken and averaged for thickness measurements.

* David.son OptronJcs Inc. Plano-lnterf erometer Model D-309 d - 5.0 RESULTS AND DISCUSSION

The primary objective of the proton experiments was to obtain sufficient data for predicting degradation during an extended period in a synchronousorbit environment, Secondary objectives of the experiments were to determine whether different substrates and coating batches affect the amount or type of radiation.damage,.and whether any physical changesoccur on the specimen. surface during irradiation. To accomplish these objectives, radiation experiments were run to determine dependence of damage onproton flux and integrated flux. All experiments were run usingambient specimen temperature and 10 keV protons.Results showed that substantial degradation in reflectance occurred in selected wavelength-bands. during. irradiatfon, Experiments and calcul.ations strongly indicated-.that the degradation-was-caused by the deposition of a proton- inducedcontaminant fiim. Inreviewing test results,the reader should be aware that-numerous effects observed during irradiation may be related to the presence of a contaminantfilm.

Typical experimental results for MgF2 and LiF overcoated mirrors are presented. first, and then are jointly discussed in a subsequent section.

5.1 MgF2/Aluminum CoatedMirrors

In this section of .the report typical: data is presented from studies onproton-integratedGf1ux dependence, flux dependence, post-irradiation reflectancechanges, and surface-finish evaluation.

5.1.1 Integrated-Flux Dependence.-The typicaleffects of protons on the spectral reflectance of MgP /a-coated surfaces were production 2 of: (1) a broadabsorption band centered at about 210 nm; (2) a narrow absorption band centered at about109.nm; and (3) a slight increase (1 to 2 percent)in reflectance in theregion near 130 nm. Reflectance data froma'MgF2/Al-coated Cer-Vit m5rror are shown in Figur s 6 and 7, illustrating .typical radiation effects at a flux of 1.4 x lof1 protons- cm-' sec'l. Figure 6 is a plot-of the specular reflectance versus wavelength, and Figure 7 shows the pereene change in specular 'reflectance ( AR/R x 100). Integratedflux is theparameter in eachplot. Negative values of AR/R represent a decrease in reflectance. 14 It canbe noted in Figure 7 thatirradiation with 2 x 10 protons- -2 cm produced a slightincrease in reflectance at wavelengthsbelow 140 nm, andinitiated growth of the longerwavelength absorption band. The long- wavelengthabsorption band peaks ar about190 nm (Figure 7) after the initialdose of radiation. Further exposure to radiation produces growthand broadening of the long-wavelength band and initiates the growth of a short-wavelengthabsorp ion band at about109 nm. Afteran exposureofprotons-cm -5 , the maximum change inreflectance of both - bands is 30,to 35 percent. The peak of the long-waveleagth absorption band. shifts from about 190 to 210 nm withincreasing radiation dose. It

15 is significant to note that the reflectance at 130 nm doesnot change after the initial 2 3ercentincrease exhibited after exposure to . 2 x 1014protons-cm- . It is alsosignificant to note that the reflectance in the wavelength region shorter than 105 nm remains higher than its pre-irradiation value even up to 10l6 protons-cm-2.

Reflectancemeasurements on the Cary-14 and IR-12 spectrophotometers indicated no significant degradation beyond about 600 nm wavelength..

A comparison of radiation damage producedon fused silica and Cer-Vit substratemirrors can be made in Figure 8. In this figure the percent change in reflectance at 210 nm was plotted versus integrated proton fluxfor two identicalmirrors of each type. Mirrors were irradiated at a fluxof-1.4 x 10l1protons-cm-2 sec-l-to an integrated flux of 1 x 10l6 protons-cm-2. As canbe seen from the small spreadin data points,excellent agreement was obtainedbetween the two Cer-Vie mirrorsand between the two fused silica mirrors.This reproducibility between similar specimens is an indication that the radiation and environmentalconditions were accuratelycontrolled, and that specimens fromthe same batchproduced similar results. The lack of agreement betweendata from the fused silica and Cer-Vit mirrors can probably be attributedto a slightdifference in MgF thicknesses. As discussed 2 later inSection 5.3.5, the spectral character of the reflectance changes is stronglyrelated to MgF thickness.Subtle differences 2 (probablyrelated to MgF thickness)can be noted in Figures 6 and 9 I1 before-irradiation"reflectance data for fused silica and Cer-Vit mirrors.Fused-silica substrate mirrors had a higherreflectance in thewavelength region from 130 to 250 nm and exhibited lower reflectance from90 to 110 nm.

5.1.2Flux Dependence.- The proton flux-dependence study was conducted to establish the validity of irradiating mirrors at fluxes lo3 to lo5 higherthan the space rate. Fluxesstudied in this work included 2 x lo9, 1.4 x 1010, 1.4 x and1.4 x 10l2 protons-cm-2 sec-l. It was hoped thatthe nearly thousand-fold spread in fluxes would providesufficient-information to conclude whether or not accelerated tests were valid.

Results showed that damage was essentially independent of flux up to 1.4 x loL1 protons-cm-2 sec-I, however, a flux of 1.4 x 10l2protons- cm-2 sec-l produced much less damage thanlower fluxes (for equivalent integratedfluxes). Thus, damagebecame deendent on flux between 1.4 x 1011 and1.4 x lo1' protons-cm-2 sec-B . This conclusion is apparent in a summary plotof data (Figure 10) for MgF /Al-coatedmirrors 2 irradiated at differentfluxes. Thepercent--change inreflectance at 210 nm wavelength* is plo.tted versus integrated flux, with flux as a parameter.With the exception of three points derived from a beryllium mirror, all data shown are for Cer-Vit substrate mirrors.

* This wavelength was chosenbecause it is near the peak of a broad absorption band and changes consistently with radiation exposure.

16 To illustrate the observed rate effects (FigureJO) , a AR/R value of -5 percent was obtained at an integrated flux of 1015 proKons-cm- 2 for a fluxof 1.4 x 1012protons-cm-2 sec-l. Forequivalent integrated fluxes,respective AR/R valuesof about -10.5 and -11 percent were obtained at fluxes of 1.4 x 1011and 1.4 x 1010protons-cm-2 sec-l. Similar res Its c n be noted for the specimen irradiated at 2 x lo9 protons-cm-' sec-B .

5.1.3 Post-1rradiat;ian- Reflectance Changes.,- The- primary purpose of this porticm of-the-study was to determine whether reflectance changes would occur as a function of time after irradiation, in both vacuum and air environments. Numerous reflectancemeasurements were made through- outthe program afcervarious periods in vacuum and/or air. Some mirrors were stored in air starting immediately after irradiation, and some were kept under vacuum for various periods of time before beginning air exposure.The shortest air exposure was about 18 minutes(defined by thelength of time the chamber pressure was abovetorr) , and the longest was 2600 hours.

Results.ofthese experiments showed that substantial reflectance changesoccurred. after irradiation. Reflectance increase (recove'red) with time in the-wavelength region above about 140 nm, anddecreased with time-at shorterwavelengths. Reflectance changes are similarfor storage in vacuum (5 x 10-7 torr) or at ambient air pressure, although the rate ofchange is greaterin air. To illustratethe post-irradiation reflectancechanges, the AR/Ra values at boththe short and long wavelength absorption peaks are plotted versus time accumulated after irradiationin Figure 11. The AR/Ra values are referencedto the "afterirradiation"reflectance (R,) ratherthan "before-irradiation" (R) values.

Data shown inthe figure represents specimens exposed to protons- -2 loL6 cm at a flux of about 1.4 x 1011 protons-em-2 see-'. Mirrorsexposed to air immediately after irradiatdon exhibited a -3-0 to -1.5 percentchange in reflectance (decrease in absolute reflectance) at the short wavelength peak in the first 18 minutes.The increase in refleetance at 21.0 w was slightly more than 1 percent in the same time period.. It was foundthat the reflectance at both short and-long wavelengths continued to change even up to 2600 hoursafter irradiaL-ion- The. respeetive A R/R values at 2600 hours are about -30 percent and +6.5 percent for the Czr-Vit substratemirrors. Results for fused silica mirrors showed slightly l.arger changes with time. -7 To evaluate effects ofstanding in vacuum (&5 x 10torr), a fused silica mirror was measured immediately after irradiation and then again after 50 hoursin vacuum. It was foundthat.the reflectance at the short-wavelength-absorption peak decreased about 12. percent, compared to 21 percentfor. specimens exposed a air forequivalent.time. No change

17 in reflectance occurred at 210. nm after 50 hour-s in vacuum.The specimen was thenexposed to air for time periodsof 13 hoursand 1012hour-s. As notedin Figure 11, the rate ofchange in reflectance of this specimen was considerablyhigher than specimens which had been keptin air forthe first 50 hours.Furthermore, the total change experienced after 1000 hours at bothwavelengths was greater than the specimenswhich were placed in air immediately after irradiation.

5.1.4Surface Finish Studies.- The objectives of the surface finish studies were todetermine: (1) the comparative microroughness of the threetypes of substrates; (2) theeffect of vacuum depositedcoatings onmicroroughness; and (3) whether or not the protons had anyphysical ef-fect on the irradiated surface.

A comparison of the microroughness of uncoatedKanigen-nickel/ beryllium,fused silica, and Cer-Vit substratescan be made in the electronmicrographs shown inFigure 12. These photomicrographs were prepared using.a standard replication process and byshadowing the replicawith germanium at an angle of 78 degreesfrom normal. On the polished nickel surface (Figure 12a), grooves andcold flow of material are visible, Theapproximate width of the largest groove is 260 m.

It is interesting to compare.the difference between the polished nickelsurface and..the fused silica and Cer-Vit surfaces.In contrast to the nickel surface., the Cer-Vit andfused silica surfaces do. not show any evidenceof scratches.or flow (smearang) of.materia1. This is presumed to be a result of the difference in. mechanism of material removal between -ductile and brittle materials. TheCer-Vit-and fused silica surfacesboth have a granularappearance, with the Cer-Vit appearing to bethe roughest of the two.The lackof resolution andmagnification prohibited measurementof the size distribution of surface irregularities.

Reflectance data from different types of coated mirror surfaces revealed-a substantially lower reflectance for beryllium-substrate mirrorsin the wavelength region below 600 nm. A comparisonof reflectance datafrom beryllium and Cer-Vit substrates is givenin Figure 13. It is presumed that the lower reflectance observed. for the beryllium substrate was a result of scattering.

Electron micrograph replicas of surfaces which were coated with Al and MgF were similar inappearance to Figures 12b and c. Thus, it was concluzedthat the coatings..did not signif-icantly change the surface roughness(at 30,000 X magnification).

Tb conclude whether proton irradiation affects the microroughness of mirror coatings, electron micrographs were made at higher resolution andmagnification than those shown inFigure 12. Higher resolution was obtained byshadowing the replica of the surface at a more oblique angle

18 ” (80 degreesfrom normal). A magnificationof 52,000 X was employed. An electron micrograph of a MgF -overcoated fused silica mirror, irradiated with 1OI6 protons-cm-2 at a 2f1uxof 1.4 protons-cm-2x sec-l, is shown in Figure 14a. The unirradiated area of the surface is shown in Figure14b. It canbe noted that the irradiated surface was covered withundulating patterns which are about 19 to 38 m- wide. No evidence of these patterns is present on the unirradiated portion of thesurface. A detailedexamination indicates considerable undulation butalways inthe same generaldirection. Similar surfaces were noted earlier in reference 7 (NASA CR-1024, Figures 18 and 19) where siliconoxide surfaces were irradiated with 16 keV protons. Two possibleexplanations of these patterns are: (1) preferentialsputtering ofthe MgF2 surface by theprotons; or (2) growth of chains of contaminant film molecules on the surface.

Irradiated MgF /Al-coated surfaces were examined with an interferometer to determlne2 whether there was anychange in surface height in the irradiated area. Selectedmirror specimens were overcoatedwith aluminumand then examined. It was found that a 4.5 nm increase in height occurredon a specimenwhich had been exposed to protons-cm-2 at a flux of 1.4 x protons-cm-2 sec-l. Conversely, no measurable increase in height was observed on a surface which had received the same integrated flux but at a higher rate (1.3 x 1012 protons-cm-2sec-I). This effect correlates with the observation discussed earlier that less degradation was produced at higher fluxes.

5.1.5Scattered-Light Data.- Scattered-lightmeasurements performed on the mirrors before and after irradiation did not reveal any significant change in scattering at wavelengthsof 253.7 nm or 500 nm. These results should be anticipated because of the small size of the irregularities formed 19 to 38 m.

A set of typical scattered-light distributions for a fused silica substrate mirror at wavelengthsof 500 and 253.7 nm, respectively, are shown in Figures 15 and 16. Theseplots represent the variation in relative power passing through the exit aperture as the image is scannedacross the aperture. The flattop on the peak of the distribution is a resultof the imagebeing smaller thanthe aperture. For these particular r‘ecords, the theoretical width of the spectral peak is about10.5 arc minutes.The calibrations in the vertical dimension representflux relative tothe specular peak. The stray light at a scattering angle of minusseven minutes is probably a result of a slight asymmetry in the alignment of the stainless steel baffle tube whichextends out from the exit aperture. At 253.7 nm the effect is less than at 500 nm because the reflectance of stainless steel is much lower at shorterwavelengths. Comparing thiswith the data for positive scattering angles to get a first estimate of the effect of this stray light, it can be seen that it has only a very small effect on the - distributionrecorded for positive angles. In each case (500 nm and 253.7 nm), the dark current was recorded so that appropriate corrections

19 couldbe made. Notethat in Figure15 (500 nm) thedark current was sufficiently low to have a negligible effect on the relative power signal. On theother hand, at.253.7 nm (Figure 16) theoutput mustbe corrected for dark current.

A comparison of scattered light distributions at the two wavelengths and for several specimens is shown in.Figure 17*. The datafor 253.7 nm'has been.eorrected.fordark current. The signffjcant observations whichcan be.made from this figure are: (1) there are no strong differences between substrates; the differences shown are not- much larger than the experimental error; (2) data at 500 nm is much more reliable than that at 253.7 nm because of-dark current corrections in the latter; and (3) despite the uncertainty associated with the 253.7 nm data, scattering at theshorter wavelength is definitely 6 to 9 times that of500 nm light.

A calculation was perforned to obtain an upper limit estimate of theeffect of scattering on. absolutereflectance measurements. The experimentally-observedscattering distributions were integrated over the solid angle which the internal PM subtends at the specimen when the former is receivingreflected light. These calculations imply that only 0.11 percent of the 500 nm light and 0.99 percentof the 253.7 nm light reach- ingthe PM is,scatteredlight. Thus,the upper limits forthe effect of an assumed 10 percent irradiation induced change in scattered light on reflectance data, are about 0.01 percent at 500 nm and 0.1 percent at 253.7 nm. Changes inreflectance observed. at thesewavelengths are therefore not due to scattering.

It is important to realize that the scattering data shown in Figures 15, 16 and.17 are affected by scattering from both the spherical mirrorand the test specimenmirror. Thus, a somewhat more realistic estimate of the scattering distribution of the test specimenwould account for the scattering distribution from the reflectometer spherical mirror.Unfortunately, there is no simple way tomeasure scattering from the spherical mirror alone while it is mounted in the reflectometer. However, if it is assumed thatthe spherical mirror is ofthe same initial quality and reflectance as the specimen, it can be shown that about 2/3 of the scattered light in the measured distribution comes from thespherical mirror and only1/3 from thespecimen. Applying this reasoning to the measurementof reflectance with the internal photomultiplier,scattering effects tend to balance out. This comes about because the separation of the internal photomultiplier from the spherical mirror in the.incident position, is verynearly equal to its separation fromthe specimen in the reflected position.

* Recorded data(Figures 15 and 16) were corrected to radiance ratios by dividing by 2.81 to correct for the size difference between entrance andscanning apertures.

20 5.2LiF/Aluminum Coated Mirrors

Typical radiation effects data on LiF/Al-coated mirrors are discussed in this section. The irradiationprocedures and experiments on these mirrorsweresimilar to that used on MgF /"coated mirrors. 2 5.2.1Integrated-Flux Dependence - Typicaleffects of proton irradiation on-the LiF/Al-coated fused silica mirrors, are shown in Figures 18 and19. Figure 18 is a plotof the specular reflectance versus wavelengthand Figure 19 shows the percentage change-in-reflectance ( AR/R), versuswavelength. Irradiation with protons resulted in production of a rather complex,broad absorption band in the wavelength region from about105 nm to 600 nm, andnarrow absorption bands at about94.5 nm and 100 nm. Thegrowth of thevarious absorption bands with increased irradiation can be noted in Figure.19.After the initial exposure incre- ment (5x 1014 protons-cm-2) , subpeaks were. apparent at 120, 145 and 190 nm. The 120and 190 nm peaks were ofcomparable heights until an exposureof 2 x 1015 protons-cm'2 was exceeded.

Thegrowth rate of peaks .at 95,120 and 200 IJ~can be. compared in Figure 20.The percentagechange-in-reflectance at thethree respective wavelengths is plottedversus the integratedproton flux. Data from two identicallyirradiated specimens were used. Test results show thatthe 95 and 200 nm-peakscontinued to grow with increasing integrated flux up to 1 x 1016 protons-cm-2,whereas, the 120 nm peakapproached saturation above 1015 protons-cm-2.

Scattered light measurements made before irradiation'and after 1 x 1016 protons-cm-2 showed .that no measurable change in scattered light occurred at either 253.7 nm. or 500 nm wavelengths.

5-2.2Flux Dependence - The validityof testing LiF/Al-coated mirrors at accelerated.rates-wasevaluated by irra iatin at fluxesof 1.2 x 1O1O, 1.3 'x 1011, and 1.2 x 1OI2 protons-cm-' see- H . The effects ofexposing specimens at different rates are shown inFigure 21.The percentage change-in-reflectance of three specimens,irradiated at different fluxes to an integrated flux of1015 protons-cm-2, is plotted versuswavelength. The most significant observation to be made from the curves is that the amount of radiation damage is both flux dependent andwavelength dependent. For example, at the 200 nm wavelengthpeak a large f lux-dependence occurred between fluxes of 1.3 x 1OI1 and 1.2 x 1OI2 protons-cm-2 see-1, with little dependence between 1.2 x 1O1O and 1.3 protons-cm'2 se6-l. On theother hand, at the 120 nm x 12 peak a small. flux-dependence was notedbetween 1.3 x 10l1 and 1.2 x 10 protons-cm-2 sec-1, and a laEgeflux-dependence between.1.2 x 1O1O and 1.3 x 1011 protons-cm-2 sec-l.

21 .

5.2.3 Post-IrradiationReflectance Changes.-The LiF/Al-coated mirrors were evaluated for post-irradiation reflectance changes similarly to the MgF2/Al-coated mirrors. Specimens were storedin ambient air, vacuum (- 5 x torr), and argon (99.996 percentlabel purity) for variousperiods of time after irradiation. It was found thatthe reflectance changed with time after irradiation and that changes occurred in both air andargon.environments. The rate ofchange in air was considerablyhigher than in argon. Thespecimen retained in vacuum after irradiation showed no significant reflectance changes after 16 hours.

Results of the experiment in which a mirror was exposed to ambient air after irradiation with 10l6 protons-nm-2 are shown in Figure 22. The percentage change-in-reflectance is.plotted versus wavelength for- conditionsimmediately after.irradTation, after 14 minutes in air, and after 38, hours in air. The 14 minute air exposure was accomplished by backfilling thechamber with air to atmospheric pressure and then immeckfatelypumping down. The time requiredfor backfilling and pumping down to 10-3 torrpressure was 14 minutes. It canbe noted in the figure that degradation continued in the wavelength region below110 nm and at wavelengthslonger than about 130 nm. A small increase in reflectance(or recovery) occurred between about 110 nm and130 nm. The width of the band in whichrecovery occurred widened by progressing to longerwavelengths as time increased.

Data from theLiF/Al-coated mirror which was allowed to stand in argon,for 62 hours .after irradiation are shown in Figure 23. It canbe noted. that the reflectance degraded .in the wavelength regions below 108. nm. and.-above .130 nm, similar to the specimenwhich was exposed to air (Figure.22). The rate ofdegradation observed in argon, however, was considerablylower than that .which occurred in air. Contraryto results obtained for exposure to air, no reflectanceincrease occurred in the,region between 108 and 130 nm.

Resultsof the above experiments showed that the magnitude of post- irradiation ref.lectance change6 is dependentupon the environment to which,.thespecimen is exposed.The question as towhether an oxygen or water.vapor reaction is involvedcannot be answered with available knowledge.

22 5.3 Discussionof Results * It was suspected early in the program that proton-induced contamination effects were the dominant cause of observed reflectance degradation. Thesuspected mechanisminvolved was radiationcross-linking or polymerization of organic molecules which impinge on the mirror surface . from the vacuum system. To verify this suspicion a ratherextensive investigation of thecontamination phenomena was undertaken.The approach involved: (1) experimentsand calculations to determine whether a contaminantfilm was depositedduring irradiation; (2) cleaning experiments todetermine the extent to which such Contaminant films affected the data; and (3) calculationsto predict the spectral reflectance changes produced-by a' contaminantfilm.

Much emphasis was placed on the investigation of contamination because such phenomena can occur in space or during environmental tests of opticalmirrors. In-the following discussion the term "contaminantfilm" is used rather loosely to explain the increase in surface height on the irradiated area. It shouldbe noted, however, that a chemicalanalysis of the film was not made and therefore the possibility exists that different-phenomenaoccurred. A discussionof results of experiments aimed at determining.whether a contaminantfilm was deposited on the mirrorsfollows:

5.3.1 ContaminantDetection Experiments.- Evldence that a contaminant film formed during irradiation came from several observations and.experiments.First, it was noted.that a water-break-freefilm would notform-on irradiated test specimens.This suggested the presence of an organicfilm. Second, anexperiment in whichchanges ininterference characteristlcs of a speciallyprepared mirror were evaluatedindicated depositionof a contaminantfilm during irradiation. The specialmirror was coated with aluminumand overcoated with vacuum-deposited fused silica (2 micronsthick) to produceadequate interference modulations on thereflectance curve throughout the visible wavelength'region*. This techniquefor detecting contamination (ref. 12) involves measuring the reflectance before irradiation, after irradiation, and after cleaning withfinely-divided calcium carbonate. The nature of changes in interference characteristics produced by irradiation and cleaning.provided information for concluding whether a contaminantfilm was deposited during irradiation.

Thecontamination deteetion speeimen was exposed toprotons- cm-2 at a flux of l0l1 protons-cm-2 sec-l. Reflectancedata before and afterirradiation and.after cleaning are shown inFigure 24. Data were plotted as straight line segmentsbetween interference maxima andminima positions. It canbe noted that the irradiation produced

* After preparation, the &.rmx was eqsese$ $0 ultraviolet radiation in air for about 18 hours ts reduce the abssrption in the ultraviolet wavelengthregion. A Genera& Eleetrkc UA3 low pressaremercury-arc lamp was used.

23 . both a shift in wavelengthof maxima and midma .et0 shorter values and a decrease in reflectance of interference minima at wavelengthsbelow about 500 nm; Themagnftude. of thechanges at midma positions became progress- sivelygreater as wavelengthsdecreased. A slightdecrease in reflectance at maxima posftions was apparentin the.region below380 nm.An estimate of thecontaminant film thicknesg, from calculations performed in ref. 12 assuming opticalconstants of n = 2 (refractiveindex) and k _= 2 (extinction coefficient)-," showed it to be less than 0.5 nm thtck. It may be recalled that interferometer.measuremenis"(Sec. 5.1.4)on a similarlyirradiated specimen indicated a 'filmthickness of 4,5 nm. This difference in thickness suggest,s that one or both of the above assumed optical constants is toolarge.

After the post-irradiation reflectance measurements had been made, the fused silica/Al-coated mirror was evaluated for water wetting characteristics. It was found that~the irradiated area would not wet with water, whereas,surrounding unirradiated areas formed a break-free film.This was a good.indicationthat an organic film had beenaffixed tothe.irradiated area. The surface wae thenscrubbed with CaCO 3 on a wet cot,tonpad. Scrubbing was continueduntil the irradiated area formed a water break-freefilm; An interesting .observation made during scrubbing was thatthe CaCO was preferentiallyattracted to the irradiated area such that ig couldnot be flushed off with water. Scrubbingwith a camel-hairbrush under running water removed the CaCO fromthat area. The implicationof this effect was thatthe irraaiated. area may have become sufficiently rough to trap the CaC03 particles. Thisobservation is substantiatedwith electron micrograph results(Figure 14).

Following cleaning, the reflectances at interference maxima and minima positions were restored to their pre-irradiation values, thus, indicatingthe removal of a thin surface film from the fused silica. Since no change in the wavelength position of maxima andminima occurred duringcleaning; it was concludedthat no fused silica was removed in thatoperation. Thepermanent shiftof interference maxima andminima wavelengthpositions is evidence of^ radiation effects in the fused silica film.

The general conclusion from this contamination detection experiment was that a very thin contaminant film was formed on the surface during irradiation.

A third- contamination detection experiment was conducted with a platinumcoated mirror. The platinumfilm was irradiatedwith protons, similarlyto other mirrors, and measured forreflectance changes. It was assumed that the optical properties of platinum would not changeduring irradiation, therefore, if a change inreflectance occurred it couldbe related.to the presence of a contaminantfilm. The platinum mirror,furnished by the Naval ResearchLaboratory (NU), comprised

24 10 nm ofplatinum deposited on plate glass. The platinum was applied usingstandard procedures. The mirror was irradiatedwith protons- at: a fluxof 1.4 x 1011 protons-cm-2 sec-l. Reflectanceand AR/R data are shown inFigures 25and 26, respectively. Data are givenin Figure 26 for various integrated fluxes, after 16 hours standing in vacuum,and afterstanding in air for 68 hours.The proton irradiation caused a rather large decrease in reflectance over the entire wavelength region (95 to 250 nm). The plotof AR/R shows thatthe decrease in reflectance reached a maximum of -4O.percent at about 120 nm wavelength. Post-irradiation reflectance measurements on the platinum coated mirror showed continueddecreases in reflectance throughout the wavelength region. It shouldbe noted that the largest post-irradiation reflectance changes occurred in the region where maximum changes occurred during irradiation.

In comparing the post-irradiation reflectance changes observed on MgF2, LiF, and platinum coated mirrors, certain differences in behavior are apparent.Results obtained on MgF andLiF overcoated mirrors 2 showed recovery in one wavelength region and continued degradation in others,whereas, platinum continued to degrade at all wavelengths measured. A possibleexplanation for these post-irradiation reflectance changes is the presence of an unstable polymer film which reacts with oxygen or water vaporover a longperiod of time. The resulting optical property changes in the contaminant film in conjunction with optical properties of the mirror coatings, could explain the difference in behavior between coating types.

A fourth experiment which indicated the presence of a contaminant film involved irradiation and subsequent cleaning of a quartz window. The quartz window.was irradiated with 10l6 protons-cm-2 at a flux of about LO1’ protons-cm-2 sec-l. Thetransmittance at 200 nm wavelength decreasedfrom 89 to82 percent (ref. 16). Subsequent cleaning with finely-divided calcium carbonate restored the transmittance to 89 percent, thus, indicating that the loss in transmittance was caused by absorption in a contaminantfilm.

25 Theabove experimen-ts stronglyindicated t-hat a contaminantfilm was appliedto the surface.dnring irradiation. Based onthe assumption that the mechanismof deposition involved polymerization of organic moleculeson the surface, further studies were conductedto attempt to verifysuch a mechanism.These studiesincluded: (1) residual gas analyses in the test chamber todetermine the gases present; (2) calculationsto predict contaminant film growth rate and an irradiation experiment to determine whether the contaminant film deposition rate is controlled by the vacuum environment or radiation parameters (flux and integratedflux); (3) cleaningexperiments to measure the residual optical damage after removal of thecontaminant films; and (4) calculationsto predict ref-lectance changes induced by the presenceof a contaminantfilm. Discussions of results of thesevarious studies are given in the following sections.

5.3.2 Residual Gas Analyses.-Residual gas analyses were made in the test chamber using a magnetic,sector instrument*. Mass spectra were obtained with the chamber blankedoff from both the accelerator and monochromator,during the reflectance measurement mode of operation, and during.the.irradiation mode ofoperation. All conditions were similar to actual mirror tests with the exception that the proton rf source could not be operated-during the irradiation mode because of electrical noise problems.

Mass analysis data for the three modes of operation were adjusted forcomparable output at atomic mass unit (amu) 14(assuming the isolated chamber c to be- baseline condition) and are plotted in Figures 27, 28 and"29.Determination of the actual gas composition and sources of contaminants was not undertaken because it was not within the scope of thepresent program. However, cursoryevaluation revealed the followinginformation:

The dominant gas species for any mode of operation was hydrogen. Operation of the system with either the monochromator or low energy accelerator (both of which inject hydrogengas) produced essentially 100 percenthydrogen in the chamber. Pumping on the chamber for three days only reduced the hydrogen concentra- tion to about 50 percent-

A relatively. high cancentration.of hydrocarbon molecules with masses up to about-90.amu.was-present in both the isolated teat chamber,and the chamber when open to the monochromat-or. Although a. silicone spectrum,-wasanticipated (due to DC-704 oil in-the diffusion pump), it was not detected;

Thehydrocarbon molecules were not effectively trapped-out with the liquid nitrogen.trap in the test chamber;

A substantial decrease in the proportion of hydrocarbonspresent " in the chamber was observed during operation with the accelerator.

* VEECO Instruments Inc. Model GA-4R

26 Inregard to sources of hydrocarbon molecules, several interesting observations can be made. It was notedthat the cylindrical cold trap a in the chamber was not effective in trapping out organic molecules. This may indicate that the source of the hydrocarbons was semi-infinite. (Backstreamingfrom a pump would be a semi-infinite source, whereas, outgassing molecules from O-rings would decrease with time as the volatile molecules were depleted.)The observation that the hydrocarbon molecules appear to come from a semi-infinite source correlates with the fact that reflectance degradation on mirror specimens was reproducible throughout the program.

The proportdona1 decrease of hydrocarbons during operation with the acceleratorindicates that: (1) theion pump (onthe accelerator) is more effective in pumping hydrocarbonsthan is the turbomolecular pump; and(2) the source of hydrocarbons in the test chamber may havebeen the vacuumpumping system.

5.3.3 ContaminantFilm Deposition Rate.- To verifythe hypothesis that a hydrocarbon film was deposited on the mirror surfaces during irradiation, a calculation was made usingkinetic gas theory. It was assumed that hydrocarbon molecules impinge on the surface at a rate (v) given by the following equation (ref. 17) :

22 -2 -1 v = 3.513 x 10 molecules-cm sec (m)1'2 where : P = pressure, mm Hg (torr) M = molecular weight 0 T = surfacetemperature, K

It was also assumed that the partial pressure of the hydrocarbon molecules was 1 x lo-' torr,the hydrocarbon species was CH4, the specimentempera- ture was 300°K, and the time required for irradiafion was-22 hours.

Results showed that if all impingingmolecules were permanently affixed to the surface by protons, a filmthickness of about 30-nm could beaccumulated in 22 hours.In this case the rate ofdeposition would becontrolled by the arrival rate ofthe molecules. If 'it is assumed that the rate of deposition is limited by the proton flux* (10" protons- cm-2 sec-') , a film thickness of about 6 & would be accumulated in 22 hours.This latter value is inreasonable agreement with the measured increase in surfaceheight of 4.5 ma (See.5.1.4), suggesting that the accumulation is primarily limited by proton flux for this case.

I- * This assumption implies that one molecule is affixed for each arriving proton. Hence, thecontaminant film thickness is proportionalto the - integratedflux. - 27 -2 On the other band,an increase of the flux to 1OI2 protons-cm sec" reducesthe exposure by a factor of ten(to 2.2 hours) for the same integratedflux of 10l6 protons-cm-2. If everyimpinging moleculeadheres to the surface, this shorter exposure permits accumulatingonly 3 nm ofcontaminant. This value is clearly less than the thickness predicted byassuming thecontaminant build-up to be proton flux limited; since accumulation is proportional to integratedflux, this prediction is - as before - 6 nm. Thus, forthe higherflux it mustbe concluded that contaminant deposition is limitad by thecontaminant partial pressure, and that the lesser accumulation will result in less reflectancedegradation. This much is confirmed by experiment(Sec. 5.1.2).

To attempt to verify this conclusion, a mirror was irradiated at thehigher flux in an atmosphere of reduced hydrocarbons. This should have resulted in reduction in contaminant accumulation and proportionally less reflectancechange. Irradiation was carriedout in the accelerator beam tube(see Figure 2) isolatedfrom the test chamber; flux was 2.3 x 1012 protons-cm'2 sec-l. Residualgas analysis data (Figures 27 and29) indicated that this vacuum environment would have a lower partialpressure of hydrocarbons.Results of the beam tubeexperiment showed that the amountand spectralcharacter of the reflectance changes were approximately the same as thoseinduced at similar flux anddosage in the test chamber.The most likelyexplanation for the inconsistency of this result is thatthe assumption of-fewer hydrocarbons in the beam tube, is invalid. In thisconnection it shouldbe noted that the residual gas analyses. were atypical of irradiation conditions in that ion gun could not be operated during the time of analysis.

5.3.4Cleaning Experiments.- Since results of contamination detectionexperiments strongly indicated that a contaminantfilm was deposit~edduring irradiation, it was imperativeto delineate the changes in reflectance caused by contaminationfrom those caused by proton damage. To accomplishthis, cleaning experiments were undertaken. It was hoped that by noting the reflectance before and after removal of- thecontaminant film, the permanent radiation damage to the reflective coatingscould be ascertained*. A discussionof the results ofclean- ingexperiments follows.

Observationsof irradiated MgF /Al-coatedmirrors prior to cleaning revealed that an unusual breath patgern (diffuse appearance caused by condensedbreath vapors) occurred in the irradiated zone. It was also noted that a water-break-freefilm would notform anywhere on the surface.

An unsuccessful attempt was made to remove the contaminant film from a MgF /Al-coatedmirror by a Freonand Collodion treatment. The 2 surface was flushedwith Freon prior to application ofCollodion. No change in reflectance occurred in the vacuum ultraviolet wavelength

* The majority of protons will pass through the reflective films regardless of thepresence of a thincontaminant film. The projected rangeof 10 keV protons in LiF, MgF2 andaluminum was calculated to be 105, 99 and 120 nm, respectively.

28 region.The same mirror was thenscrubbed with finely-divided CaCO using a wet surgicalcotton pad.The calcium carbonate was carefudy decantedto obtain only the smallest particlesfor scrubbing. After several light passes with the cotton it was noted that the surface wet better, however, theunusual breath pattern was still present in the irradiatedzone. The scrubbing was thencontinued until fine scratches appearedon the mirror. The unusual breath pattern could still be produced inthe irradiated zone after cleaning.Reflectance data after cleaning is shown inFigure 30. Theeffect of the abrasive cleaning was to: (1) eliminatethe strong absorption bands at 108 and 210 nm; (2) increase the reflectance above-pre-irradiation values in the wavelength region below 116 nm; and (3) only partially restore the reflectance in the wavelength region from 160 to 350 nm (the limit of measurement)*.These latter changesindicate that the abrasive cleaning may havereduced. the thickness of the MgF film.Visual. observations 2 indicate that the scattered light from the. surface should have increased,however, the effect of sueh.losses on specularreflectance was notdetermined. It was generallyconcluded that the calcium carbonatecleaning technique was tooharsh.

An attempt was made to remove the contaminant film by soaking in carbontetrachloride. It was found.that no change inreflectance occurred after soakingfor 11 hoursat-ambient temperature. These results are consistent with those of other experimenters-in attempting to remove radiation-polymerizedorganic films (refs. 7, 18, 19 and 20).

It was theorized that if the contaminant film was hydrocarbon(in contrast to silicon), it might be possible to remove it by oxidation with fluorineor atomic oxygen.Oxidation should produce gaseous or volatile compounds. It was anticipatedthat the MgF orLiF overcoatings would 2 not be decomposed because they already are stable fluorine compounds.

Exposureof pieces of contaminated aluminumand stainless steel to fluorine was attemptedfirst. These metals hadbeen irradiated over about a six monthperiod and were discolored to a yellow-brownappearance. Thespecimens were placed in small bell-jars which were evacuated and purged twice withdry nitrogen before introducingthe fluorine. This minimized the possibility of forming hydrofluoric acid from residual water vapor.The bell jars were then filled with fluorine to a pressure ofabout one atmosphere. It was notedthat the fluorine caused the yellowfilm to disappear over about a 45 minuteperiod. A similar fluorine exposure was then given to an irradiated, MgF /Al-coatedmirror 2 exceptthat the exposure time was reducedto 35 minutes. It was noted that the entire mirror surface wet with water after exposure to fluorine, but that the unusual breath pattern was still visible ;tn the irradiated area. Subsequentreflectance measurements showed thatsubstantial changes.in reflectance had occurred in the vacuum-ultraviolet wavelength region.

*The ultraviolet reflectometer was initially designed for operation at wavelengthsshorter than 250 nm. It was demonstrated later thatreason- ably accurate measurements could be made out to wavelengths of 350 to 400 nm.

29 Specular reflectance and AR/R data are shown in Figures 31 and32, respectively. It canbe noted in Figure 32 thatthe fluorine nearly eliminatedthe absorption band centered at 210 nm, andincreased the absorptionin the wavelength region below113 nm. Examinationof this mirrorwith an interferometer showed that a 5.9 nm thick film was still presenton the irradiated area. Thissuggests that the fluorine may not have removed thecontaminant film, but only changed its optical propert-ies.Three other irradiated. MgF2/Al-coated mirrors were also exposed tofluorine with varying results. In all cases thefluorine nearlyeliminated the absorption band centered at 210 nm, however,below about130 nm data was conflicting. It was notedthat two of the MgF2/A1- coatedmirrors turned slightly hazy (or diffuse) in the irradiated area as a result of exposure to fluorine.

An unirradiated MgF /Al-coated .mirror was exposed to fluorine to ascertain whether reflecgance changes discussed above, coul~d have been causedby changes .in the MgF2 or aluminum films. It was found that the reflectanceincreased in the wavelength region from 106 to180 nm. The maximum increase noted was 9 percent ( AR/R x 100) at 110 nm (Figure 33). Possible explanations for this increase in reflectance of the unirradiated mirrorinclude: (1) optical property changes in the MgF2 filmresulting from elimination of-an inherentfluorine deficiency; and(2) a thickness changeof the MgF film. Ifan inherent def-iciency does exist , the results suggest tiat mirrors of this type should be treated with fluorine af t-er coating . Lithium fiuoride overcoated mirrors were also exposed to fluorine in an attempt to remove thecontaminant film by oxidation.The results were completely unsatisfactory because of etching.

In genexal,the fluorine cleaxing technique.did not provide satisfactoryresults,,therefore, experiments were Lnitiatedto evaluate atomfcoxygen'as a cleaningagent. The oxygen cleaning technique involved exposingmirror specimens in an atomic oxygen plasma created by an rf~ source.The mirror specimen was placedin a vacuumchamber which was evacuatedto about 5 x torrpressure. Thechamber was then back- filled with oxygen to a pressure of 4.5 x 10-1 torr. An rfantenna, operated at one endof the chamber,provided sufficient excitation energyto ionize oxygenthroughout the chamber. No acceleratingpotentials were applied in .the chamber.

Results ofoxygen cleaning an Srrad.iated MgF /Al-coatedmirror are 2 givenin Figure 34.The mirror was exposed foronly 5 minutes.The reflectance at wavelengthsabove about 140 nm was restored to within 2 percentof its valuebefore irradiation. In the region below 120 nm the reflectance exceeded the pre-irradiation value by as much as 20 percent ( AR/R x 100). NQ visible effects were presenton the mirror surface.Interferometric examination of the surface showed thatthe contaminant film was not present after cleaning.

30 Theoxygen cleaning technique was then evaluated on an irradiated LiF/Al-coatedmirror. Results are shown in Figures35 and 36. It was found that a 5 minute exposure restored the reflectance to within 7 percentof its pre-irradiationvalue at all wavelengths. An additional 5 minuteexposure, however, began to degrade the reflectance again in the region from 99 nm to 112 nm and initiated crazing on the mirrorsurface in one small patch.This result suggests that a 5 minute exposure is sufficientlylong, or too long for removing the contaminant film. Interferometricexamination revealed that the contaminant film was not present after cleaning.

In general, the atomic oxygen cleaning technique was highly successful in demonstrating that the bulk of the damage observed during irradiation was caused by thecontaminant film. The small net change whichremained after cleaning could be the result of proton damage to the reflecting coatings, the result of residual contamination, or theresult of removing a small amount o€ the MgF coating.Analysis 2 on a theoretical basis (as in Section5.3.5) favors interpretation as removal of a small amount of MgF coating. A cleaningtechnique of this 2 type may be of great value for both commercial and space applications. Forexample, in commercialapplications it may be useful for cleaning hydrocarbonsfrom optical mirrors and prisms, and from surfaces which are tobe prepared for bonding. A possiblespace-program application is restoring the reflectance or transmittance of optics which have been contaminatedin space or during Earth tests. Sincecontamination of this type is a significantproblem facing space optical systems, it is recommended that the oxygen technique be developed and evaluated for such uses.

5.3.5Prediction of Contaminant Film Induced Reflectance Changes.-. Calculation of reflectance values from classical electromagnetic theory forms the basis for .an exacting test of the hypothesis that mirror degradationresulted from the deposition of. a contaminantfilm. For lightof zero polarization, incident normally, the calculation requires only a knowledgeof the thicknesses and optical constants (refractive index, n, and extinctioncoefficient, k) of the various layers which make up thereflecting boundary. These conditions on theincident light are sufficientlyapproximated in the experiment (7.5 deg. of arc incidence in monochromator,6.0 in reflectometer) so as toinvolve negligible error in comparing measured and calcqlated reflectances.

For the work describedbelow, solutions of the electromagnetic equations, based on a transmission line modelof the boundary value problem, were programmed for a digital computer.Reflectances were calculated at 2.5 nm intervals from 95 to 345 nm, for a structure of opaquealuminum overcoated with magnesium fluoride? A contaminant layer was thenadded and reflectances were recalculated. It was desiredthat thecalculations predict: (1) reflectances before and afterproton irradiation,that is withand without the contaminant film; (2) changes in the damage spectrum (AR/R vs A ) as thedosage (or contaminant film thickness) is increased;and (3) changes which result when a mirrorwith - increasedthickness of magnesium fluoride is irradiated.

See Appendix foroptical constants for MgF andaluminum. * C 2 31 A basic difficulty encountered was the lack of optical constants datafor the contaminant film. Following a suggestion(ref. 21) that maximum changeof reflectance should occur at interference minimaand minimum change at interference maxima, a series ofpreliminary calculations were performed* to relate reflectance to optical constants and thicknessof a contaminantfilm at selectedwavelengths (see Appendix D foradditional detail). These results were compared withtypical observedreflectance changes and contaminant film thickness data to obtain n and k values at two wavelengths. By drawingdispersion-like curvesthrough each pair of points, these data were extrapolated to other wavelengths.

With thissimple, first approximation, the spectral dependence of reflectance was exploredthroughout the wavelength region showing large changesduring radiation. The success achieved shows that all major changesobserved in the test program are explainable in terms of contamination. The salientfeatures of thepredicted changes - all confirmed by relation to observed effects - are the following:

1) For themirrors coated with 25 nm of-magnesium fluoride,the change in reflectance occurs in two degradationpeaks - a relatively narrow band centered near 110 nm and a broad band near 220 nm; 2) Between thesedegradation peaks the change in reflectance drops effectively to zero at 130 nm; 3) With increasingdosage, each of the three main features described above shifts to longer wavelength andshows increased degradation; 4) For a mirrorcoated with 43 nm of magnesium fluoride,these same threefeatures appear, but shifted to longer wavelengths.

Thebehavior described in (1), (2) and (3) canbe observed in the curves of Figure 37, which show degradation as a function of wavelength forseveral thicknesses ofcontaminant. The agreement between predicted andobserved degradation is shown in Figure 38.

The experimental data for a fused silica mirrorserved as input datafor determining the contaminant optical constants (Appendix D). As a result, exact agreement was expectedbetween the experimental- degradation at 200 nm and that calculated for a thickcontaminant layer.Approximate agreement was expected at 121.6 nm. Agreement was realized within the consistency of the optical constants for aluminum andmagnesium fluoride used in the preliminary calculations and the presentcalculations. Theagreement at otherwavelengths is, of course, thesubstance of the argument for the contaminant theory. The excellent agreementof the short-wavelength degradation peak and the intermediate

* By Mr. W. R. Hunterof the Naval Research Laboratory -

minimum at about 130 nm with experimental results, should be particularly notedin Figure 38. At wavelengthsshorter than 110 nm, agreement betweencalculated and measured values is less convincingthan at longer wavelengths.This is largely a resultof uncertainty in the optical constants of magnesium fluoride which is resonant at wavelengths near 110 nm. The experimentalcurve of Figure 38 in thisregion has been drawn in consonance with the calculated shape, but it should be noted that the density ofexperimental points is insufficient to render the detail shown.

To test the theory for a case of a thickercontaminant film, calculated reflectance changes were compared to data from a mirror which was irradiatedwith more than1016 protons-cm-2*(Specimen T-19). It was foundthat changes produced by a 15 nm-thickcontaminant film, provided thebest fit to experimental data (Figure 39). A significantobservation to be made in the figure is the ability of the theory to predict the shift of maxima and minima positions to longer wavelengths (compared to those shown inFigure 38). Althoughcalculated data shown inFigure 39, assumgs a 15 nm-thickc0ntaminan.t film, interferometric fglm-thickness measurements**and the maximum degradation experienced at the long- wavelengthabsorption peak, suggest a slightly thinner film.

The calculated effect of depositing a 5 nm-thickcontaminant film on a mirrorovercoated with 43 nm of MgF is shown inFigure 40. Experi- 2 ***, mentaldata for an exposure of 1016 protons-cm-2 are also shown. In comparingexperimental and predicted damage curves,reasonable agreement is seen between the order and magnitude of the maxima andminima. However, all features-on the theoretical curve occur at longerwavelengths thantheir experimental counterpart. Here again,analysis of a trend provesusgful. The change in predicted and measured behavior caused by increasing the magnesium fluoride coating thickness from 2.5 m to 43 nm canbe observed by correlating Figures 38 and 40. As can beseen, thetrend of change predicted by calculation is confirmed in experiment.

?hebasic agreement between calculated and measured reflectance is shown inFigures 41 and 42. Figure 41 compares thecalculated reflectance for a 25 nm MgF2 coating with the reflectance values obtained for the specimens of Figures 38 and 39 before proton irradiation. Figure 42 provides similar information for the mirror coatzd with 43 nm of MgF2 whosedamage spectrum is shown in Figure 40. Forboth MgF2 thicknesses,the agreement is generally good forlong wavelengths (> 140 nm), fair at the onset of MgF2 absorption (110 to 140 nm), and rather indifferent at wavelengths shorter than the absorption edge ( <110 nm).

* Thismirror was usedfor checkout of the system and received an unknown integratedflux above protons-cm-2. - ** An area weighted mean of 13.6 nm *** Thismirror was provided by Dr. J. F. Osantowski,Goddard Space Flight Center; the 43 nmthicknessquotation was supplied with the mirror.

33 Analysisof the residual disagreement between theory and experiment- is noteasy. A partialanalysis indicates that better agreement would beobtained for the thick MgF coatingwith an assumed MgF thicknessabout 10 percent less. Slmilarly,2 an increase of 5 to10 2 percent in the assumed 25 nm MgF coatingthickness wouldimprove agreement 2 of the basic reflectance curve at theabsorption edge and could improve the damage spectrum match.

At wavelengthsshorter than 110 nm, betteragreement with experiment might be achieved by tailoring the optical constants of magnesium fluoride. It is notimmediately apparent that a uniquedescription suitable for all specimensof magnesium fluoride could be achieved; that Ps, each crystal fromwhich coatings are preparedmight have slightly varied absorption characteristics.

5.3.6Degradation In Space.-The primary objective of this study was to determine the effects of thesynchronous-orbit proton environment onmirror surfaces. This section is thereforedevoted to discussing program resultswith respect to that objective. As developed in the foregoing sections, the primary mechanismof- damage was proton-induced depositionof a contaminantfilm onto the mirror surfaces. Results of cleaningexperiments and theoreticalpredictions of reflectance changes on MgF2/Al-coated mirrors, quite conclusively show that the contaminant filmaccounted for essentially all of theobserved damage. Furthermore, sincethe contaminant film was sufficiently thin (w 5 nm the irradiation ofrmirrorcoatings prop-er was notinhibited. Thus, the normalproton radiationeffects such as colorcenter production, sputtering, and blistering are apparently of negligible consequence for these surfaces and radiationexposures protons-cm-2)*.The same conclusion cannot yet.be made for LiF/Al-coated mirrors because reflectance changes havenot been theoretically predicted. Cleaning experiments on those mirrorsindicated, however, thatthe majority ordamage could be accounted tothe contaminant film.

It may be recalled from Section 2.0 that at synchronous orbit a mirror surface with ZIT geometry will encounterabout 5.5 x 1015protons- year'l,with roughly 70 percenthaving energies less than 30 keV. Assuming no dependence of damage onproton energy, (i.e., 10 keV protons adequatelysimulate the energy spectrum in space), an integrated flux of

1016protons-cm-2 representsabout two yearsin synchronous orbit; ~ The overall program results thus lead to the conclusion that negligible degradation will occuron MgF/Al-coated mirrors over a two year period in synchronousorbit. The same conclusioncan probably be stated for LiF/Al-coated mirrors, with the reservation that contaminant-film induced effects have not yet been verified by theory.

* A cautionary note in this regard is consideration that the presence of a contaminant film may have interfered with mechanisms for loss ofatoms from the fluoride coating during irradiation.

34 6.0 CONCLUSIONS AND EBCOPIPIENDATIONS

As a result of experiments conducted in this program, the following conclusions have been reached:

Protonexposure, equivalent to about two-years of unshielded operation at synchronousorbit, induced no significant reflectance change in either MgF2/Al or LiF/Al-coatings in the wavelengthinterval of 90 to 50,000 nm;

Mirrors coated with MgF2/Al andLiF/Al (as well as other coatings) exhibit significant reflectance loss in the vacuum ultraviolet as a result of proton induced deposition of a contaminantfilm;

The reflectance of irradiated mirrors could be restored to nearly pre-irradiation values byremoving the contaminant film in an atomic oxygen plasma; -2 Irradiationmirrorsof with protons-cmcaused no measurablechange in scattered light at either 253.7 or 500 nm wavelengths;

The necessity of in-situ reflectance measurements was not established since the mirror degradation was causedby contaminant film formation rather than by coating or substrate damage ;

The typeof mirror substrate (fused silica, Cer-Vit, or Kanigen-nickelplated beryllium) was immaterial to the radiation effects observed in the experiments;

Substrates of fused silica and Cer-Vit were polished to a better finish than the Kanigen-nickel surface;

Electronphotomicrographs showed that application of vacuum depositedcoatings to polished substrates did not signifi- cantly affect the microroughness of the surface.

35 The followingrecommendations are presented as a result of this research:

1) Additionalproton radiation experiments should be conducted in ultra-high vacuum to study radiation effects in a cleaner environment.Such experiments are requiredto verify that radiation-inducedcontaminant deposition does not inhibit the coating damage mechanisms;

2) A feasibility/developmentstudy should be conducted on the oxygencleaning technique to evaluate its use for restoring reflectance of mirrors contaminated in space or in large vacuum chambers;

3) A spaceflight experiment should be developed to measure degradation of ultraviolet-reflecting mirrors down to 100 nm wavelength.Since optical properties ofthese mirrors are extremelysensitive to thin contaminant films, the experimentshould include exposures to both a "clean" and 11 outgassing-organic"environment;

4) Scatteringexperiments should be conducted with self-imaging test specimens so that absolute data can be obtained at wavelengths shorter than 253.7 nm;

5) A programshould be conducted to determine the contaminant film deposition mechanismand to develop a suitable theory for predictingcontaminant film growth from known environmental conditions. 7.0 APPENDIXES

APPENDIX A MIRROR SAMPLEPROCUREMENT SPECIFICATIONS

Mirror Application.- These mirrors (of ultraviolet astronomical telescope quality) are to be used to determine the effects of charged particle irradiation on their reflectance and specularity in the far- ultraviolet to middle-infrared wavelength region. MirrorSpecification.- Size All samplesshall be of dimensions: Lengthofsides 2.00 2 0.05 inchessquare (5.08 cm) Thickness0.75 (1.9 inch2 0.10 cm) Substrate Material Beryllium-Kanigen - Grade HP-40 from Beryllium Corporation, Reading,Pennsylvania. All surfacesto be plated with Kanigen Nickelto a thicknessof 0.006 inch (0.015 cm). Platingshall be continuouson all surfaces. Fused Silica - CorningFused Silica /I7940 Mirror Blank Quality. Cer-Vit - Premium GradeMirror Blank Material C-101, Owens- Illinois, Toledo , Ohio. Surface Figure One of the 2- by2-inch surfaces of each sample shall be polished flat over a 1.875-inch-diameter(4.76 cm) aperture to X/lO for the mercury green line (546.1 nm).

Surf ace Quality There shall be no evidence of gray or orange peel on the uncoated and finished mirror surface when viewed with a 5x eyeloupe. Surface defects classified as pits,digs, or sleeks shall not, exceed a 60-40 finish as generally defined by Mil Spec MIL-0-138308. Reflective Coatings MgF2/Al :

Coatings were applied by Optical Coating Laboratory, Inc., SantaRosa, California. Samples to be coated shall be aluminizedto a thicknessof 60 to 120 nm (just visibly opaque) with fast evaporated, high purity Al andovercoated withapproximately 25 nm of MgF 2 toproduce a minimum reflectivity of 0.73 ht 121.6 m and a reflectivity versus wavelengthcurve similar to that obtained by Hass & Tousey (J. Opt. Am. 49,601, Figure 17).

37 APPENDIX A (Contined) LiF/Al:

Coatings were applied by Dr. George Hass, U. S. Army Electronics Command, Might VisionLaboratory. Mirror samples shall be coated to a thicknessof 60 to 120 nm with fast- evaporated, high-purity aluminumand overcoatedwith approximately 14 nm ofLiF. A minimum nominal reflectivity of^ 72 percent at 102.6 nm shallbe required. The mirrortemperature shall be maintained at near ambient temperature during vacuum operations.

Uniformity

All mirrorspecimens of a given type shall be coated in the same batch.Uniformity of reflectance shall be within t 0.05 at 102.6 nm and 121.6 nm for LiF and MgF,, overcoated mirrors respectively.

A witness blank shall be included with the mirror batch during thecoating process. A strip 0.08 to 0.13 cm wide shallbe masked acrossthe center of the witness blank. Commercial microscope slides are acceptable.

Supports for the mirrors during evaporation must not--~intrude more than 0.15 cm onto the mirror face.

AlLsurfaces except the mirror surface shall be ground flat with an 80-grit finish.

Surfaces should be square with respect- to each other and the mirrorface to within 0.25 deg. of-arc,

Each sample shall be peraianently markedon theedge (vacuum proof) with a codeshowing material andsample number.

All sharpedges shall be removedby 45-deg. of arc beveling, 0.02 to 0.05 cm on the flat.

The substrate material shall be ordered from a specific and identifiablebatch or lot; Mirror samples shallbe accompanied by detailedinformation, including source and composition of material and theprocesses involved,so that duplication of procurement is possible.For example, data on vacuum coating conditions is desired.

Handling precautions a) No contact on faceof mirror after evaporation, except in support areas identified in Item 8.

38 APPENDIX A (Continued) b) No partof mirror to behandled without gloves, or other- wise contaminated with hydrocarbons after precoating cleaning. c) The mirrorsamples shall be kept in an environment of relativehumidity lower than 40 percent.Coated mirrors shall be stored andshipped in Boeing-suppliedsealed containers.

39 APPENDIX B

MISCELLANEOUS DATA ON MIRROR SUBSTRATES AND COATINGS

MgF2 andLiF Overcoated Test Specimens

Substrate Data.-

B er.yllium

Manufacturer:BerylliumCorporation The (Berylco) Specification: HP-40 BerylcoUnit Nos. : 669K-1 through 16 BerylcoFilm No. : HSB-2684 Heat No. : 669K Purity Analysis: Beryllium 95.10 percent B e0 6.68 Carbon 0.164 Iron 0.167 Aluminum 0.060 Magnesium 0.010 Silicon 0.061 Other Metallic 0.10 max. 8 -2 Ultimate Tensile Strength: 4.96 x-10 newtons-m 7 -2 Precision Elastic Limit: 6.2 x 10 newtons-m Grain Size: Less than15 microns (15000 m) 3 Density : 1.880 x 10 kg-m-3 Kanigen-NickelPlater: Grunwald Plating Company Chicago,Illinois Plat-ing Specifications : MIL-C-26074A-C1 Heat- Treatment : 463 2 5OK for 4 hours Nickel Plating Thickness: 0.015 cm (0.006in.) APPENDIX B (Continued) Cer-Vit

Manufacturer : Owens-Illinois Company, Toledo , Ohio Specification: Premium Grade Mirror Blank Material , c-101 Chemical Analysis: Meets Patent Requirements Average Linear Thermal -7 0 Expansion Coefficient: 0.0 x 10 / C at 0' to 38OC Dimensions : 5.039-5.080by 5.039-5.080 by1.864-1.905 cm Seed Count : 0 to 4 perspecimen Stress Analysis: Stress retardation of 5 - 10 mp/cm inthe diffuse stress. Stress is mainly at sharp corners.

Fused Silica

Manufacturer : Corning Glass Works, Bradford,Pa. Specification: f7940(Mirror Blank Quality) Code No. 851056) Corning Glass Order No.: (MgF OvercoatedMirrors) OZ 813326 December 28,1967 2 Corning Glass Order No.: (LiF Overcoated Mirrors) OZ 813788March 6, 1968

Cleaningand Coating Data.-

/Aluminum Coatings =2 1) Cleaned in a liquid detergent/water-solution; 2) Wiped dry with flannelcloth; 3) Glow dischargecleaned in vacuum chamber; 4) Vacuum deposited aluminum filmapplied in accordance with datapublished by Hass andTousey (J. Opt. SOC. Am., Vol.49, 593,1959) Chamber pressure ------less thantorr Rate ...... approximately 85 nm in less than4 seconds

41 APPENDIX B (Continued) 5) Magnesium fluoridefilm applied as follows:

Substrate temperature ------ambient

Chamber pressure ------less than loW5 torr Rate ...... 2'5 nm in less than 10 seconds MgF2 source ------procuredper Mil Spec JAN'"-621

(Berylliumand Cer-Vit substrates coated in OCLI coating run No. 352-163 February28, 1968. Fused silica substratescoated in batch 352-166, February29, 1968.) LiF/Aluminum Coatings 1) Glow dischargecleaned for 3 min; 2) Aluminum deposition Chamber pressure ------m 1 x IO-~ torr Time period ------2 to 3 sees.

Thickness ---".------M 80 nm

3)LiF deposition Chamber pressure ------w 2 x IO-~ torr Time period ------10 secs. Thickness ------13.5to 14 nm (LiF source was random cuttings from Harshaw crystals. Evaporatedfrom a tungsten boat.) APPENDIX c OFTICAL CONSTANTS OF ALUMINUM Refractive Ektinction Wavelength, Index, Coefficient, nanometers n k

90 -0 0.0630 0 397 95 -0 0.0544 0 535 100 .o 0.0526 0 9653 105.0 0.0526 0 757 uo.o 0 -0537 0.852 115.o 0 -0555 0.940 3.20 00 0.0578 1.026 125 -0 0.0605 1.105 130.0 0.0635 1.186 135.0 0 -0667 1.262 140 -0 0.0703 1.338 145.0 0.0740 1.410 1% 00 0 0779 1.484 155 -0 0 -0823 1. 555 x60 00 0.0863 1.625 3.65 .o 0 .OW7 1.694 3.70 90 0.0954 1= 763 175.0 O.lOO0 1.830 180 .Q 0.1050 1.89 185 -0 0 .109 1.964 190 00 0,1150 2 -031 195 -0 0.l21X) 2.096 200 00 0 .I260 2.162 205 .o o .1320 2 .227 210 .o 0.1380 2 .291 215 -0 0.1_440 2.356 220 .o 0 .1400 2.360 240 .o 0 .la0 2.600 253 -6 0.1800 2 -770 260 00 0.1900 2.850

280 .o 0.2200 3 9 130

300 .o 0 02500 3 9 330 320 .o 0 .2800 3.5a 34O.o 0.3100 3.800 360 .o 0.34OO 4.010 380 .o 0 -3700 4.250 400.0 0.4000 4.450 436 .O 0.4700 4.840 450 -0 0.5100 5 .om 492.0 0.6400 5.500 546 -0 0.8200 5.990 578 .o 0 9300 6 9 330 650 -0 1.3000 7.110

43 APPENDIX C (continued) OPTICAL CONSTANTS OF MAGNESIUM FLUORIDE Refractive Extinction Wavelength, Index, Coefficient, nanometers n k

90 -0 1.5600 0.220 92.5 1.5600 0.265 95 *o 1.5600 0 .320 97.5 1.5600 0 372 100 .o 1. 3800 0 .414 102.5 1. 4300 0.430 105.0 1.9000 0.420 106.0 2.1000 0.402 107.o 2.2200 0 -372 108-0 2 -2300 0.318 109.0 2.2200 0.262 llo00 2 .2000 0,210 111.0 2.0880 0 177 112 00 2.0220 0.152 113.0 1 9730 0 -134 114.O 1.9310 0.120 115.o 1.8940 0 .lo7 116.0 1. 8610 0 -095 117.o 1.8320 0 .OS4 118 .o 1.8050 0 0073 119.0 1.7810 0.065 120 .o 1.7593 0.055 121.0 1.7bO 0 .Ob7 122 .o 1. 7250 0,043 123-0 1.7llO 0 .Oh1 124.O 1.699 0 -039 125.o 1.689 0 -038 126.0 1.6810 0 0037 128.0 1,6660 0.036 130 .o 1.6530 0 -035 132 .o 1.6420 0.034 134 .O 1.6320 0 0033 136.o 1. 6210 0.032 138 .o 1.6120 0.031 14.0 00 1. 6030 o .030 142 .O 1.5950 0 -029 144.o 1.5860 o ,028 146 00 1 5780 0.027 148 .O 1.5670 0 -026 150 .o 1 5540 0.025 19.o 1-5390 0.024 3.54 .O 3.. 5230 o ,023 156.o 1*5070 0.022 158.o 1. 4930 0.021

44 APPERDIX C (continued) OPTICAL CONSTANTS OF MaGrNESIUM FLUOFtI.DE R efractive &tinction Refractive Wavelength, Index, Coefficient, nanometers n k 160 .o 1.4820 0 -020 162 00 0.01g 1.4750 164 00 1.4720 0 -018 166 .o 1.47-00 0 -017 168 00 1.4700 0 -016 170 0 1.4680 0.015 174 .O 1. 45%) 0 -013 178.0 1.4530 0.011 182 .o 1.4480 0 0009 186.0 1.4440 0.007 19 00 0.005 1.4420 145 -0 1.4400 0.004 200 .0 1.4390 0 -002 404.6 1.3w 0 00 589 .4 1.3895 0 00 706 5 1 3877 0 00

ASSUMED OFTICAL CONSTANTSFOR COMTAMINAm F'ILM Re fractive Extinction Refractive Wavelength, Index, Coefficient, nanometers n k

90 -0 1 570 0 -605 100 00 1.540 0 570 125 .o 1.495 0 -492 150 00 1.455 0 .430 175 -0 0.380 1.432 200 .o 0.335 1.413 225 -0 1.399 0 0295 250 .o 1.385 o ,270 275.0 1.375 275.0 0.240 300 .o 1.362 0 -220 325.0 1.355 0 195 .-1.350 350.0 0 -170

45 APPENDIX D

OPTICALCONSTANTS FOR CONTAMINANT FILM

To demonstrate the effect of a contaminantfilm on a MgF2/Al-coated reflective surface Mr. W. R. Hunterof the Naval ResearchLaboratory calculated the reflectance at 220 nm of a multilayercomprising opaque aluminum overcoated with 25 nm of magnesium fluoride plus the contaminant film.The calculations were carriedout-as a function of contaminant filmthickness from zero to 10 nm. Drawing onprevious experience, Mr. Hunterchose a value of 1.4 for the refractive index and carried out calculations for five values of extinction coefficientfrom 0.2 to 0.6. The curvesgenerated are shown in Figure43. -2 Followingirradiation protons-cmby , a MgF2/Al-coated fused silica specimen showed a degradation, AR/R = 0.333, at 220 nm. Interferometric measurement showed a mean accretion of 4.1 nm in the irradiatedzone. As shown inFigure 43, these data are-compatible with anextinction coefficient k = 0.3. Similar,but less detailedcalcu- lations byHunter at 121.6 nm showed that the assumption of a refractive index n = 1.5 and an extinction coefficient k = 0.5, while not in good agreement with experimental data, were a fair approximation.

With these two pairs of points, dispersion like curves were con- structedfor n and k as functionsof wavelength. These curves are shown in Figure 44, and data taken for the curves for calculation purpose are shown in the final table in Appendix C.

46 8.0 REFE-NCES

1. A SystemStudy of a Manned Orbital Telescope - Synchronous OrbitStudy, NASA CR-66102, 1966.

2. . A SystemStudy of a Manned OrbitalTelescope, NASA CR-66047, 1965. 3. Frank, L. A.: SeveralObservations of Low-Energy Protonsand Electronsin the Earth's Magnetospherewith OGO-3. J. Geophys Res., vol.72, 1967, pp.1905-1916.

4. Frank, L. 8.: On the Extraterrestrial RingCurrent During GeomagneticStorms. J. Geophys. Res., vol. 72,1967, pp.3753-3767.

5. Katz, L., et.al.: Low-Energy Particles Measuredby OV1-9 (1966-118). Earth'sParticles andFields, B. W. McCormac, ed., ReinholdInc., 1968, p. 103.

6. Vette, J. I., andLucero, A. B.: Models ofthe Trapped Radiation Environment, Volume 111: Electrons at SynchronousAltitude. NASA SP-3024, 1967.

7. Gillette, Roger B. : Proton and Ultraviolet Radiation Effects on SolarMirror Reflective Surfaces. J. SpacecraftRockets, vol. 5, no.4, April 1968, pp. 454-460.More detailedresults were given in NASA CR-1024, Ultraviolet-Proton Radiation Effects on Solar Concentrator Reflective Surfaces, May 1968.

8. Fogdall,Lawrence B.; Cannaday, Sheridan S.; and Brown, Richard R.: In-Situ Electron, Proton, and Ultraviolet Radiation Effects on ThermalControl Coatings. Final report on NASA/Goddard contract NAS 5-9650, January1969.

9. Hass, G.; Kamsey, J. B.; Heaney, J. B.; andTriolo, J. J.: Reflectance,Solar Absorptivity, and Thermal EmissiGity of Si0 2 - Coated Aluminum. AppliedOptics, vol. 8, no. 2, February1969.

10.

11. Reichard,Penelope J.; Triolo, Jack J.: Preflight Testing of the ATS-1 ThermalCoatings Experiment. Thermophysics of Spacecraft and PlanetaryBodies. G. B. Heller, ed., Academic Press, 1967, pp.491-513.

47 12. Hass, G.; Ramsey, J.P.;Triolo, J. J.; andAlbright, H. T.: Solar Absorptanceand Thermal Emittance of Aluminum Coated with Surface Filmsof Evaporated Aluminum Oxide.Thermophysics and Temperature Control of Spacecraftand Entry Vehicles. G. B. Heller, ed., Academic Press1966, pp. 47-60.

13. Gillette, Roger B.; Brown, Richard R.; Seiler,Richard F.; and Sheldon, W. R.: Effects of Protonsand Alpha Particleson Thermal Properties ofSpacecraft and SolarConcentrator Coatings. Thermophysicsand Temperature Control of Spacecraftand Entry Vehicles, G. B. Heller, ed., Academic Press,1966, pp. 413-440.

14. Canfield, L. R.; Has's, G.; andWaylonis, J. E. : FurtherStudies on MgF2 Overcoated Aluminum Mirrors with Highest Reflectance in the Vacuum Ultraviolet. Appl.Opt., vol. 5, no. 1, January1966, pp. 45-50.

15. Cox, J. T.; Hass-, G.; andWaylonis, J. E.: FurtherStudies on LiF-Overcoated Aluminum Mirrors with Highest Reflectance In The Vacuum Ultraviolet. Appl.Opt. , vol. 7 , no. 8, August-1968 ,. pp.1535-1539.

16. Personalcommunication with Dr. Georg Hass of the U. S. Army Electronics Command, NightVision Laboratory, Fort Belvoir, Va., January 1969.

17. Dushman, S., and Lafferty, J.: ScientificFoundations of Vacuum Technique,Wiley and Sons, New York,1962.

18, Ennos, A. E.: The Originof Specimen Contamination in the Electron Microscope. Brit. J. Appl.Phys., vol. 4, April1963, pp. 101-106.

19. Hillier, J.: On theInvestigation ofspecimen Contamination in the ElectronMicroscope. J. Appl.Phys., vol. 19, March 1948,pp. 226-230.

20. Ennos, A. E.: TheSources of Electron-Induced Contamination in Kinetic Vacuum Systems. Brit. J. Appl.Phys., vol. 5, January1954, pp. 27-31.

21. Personal Communication with D'r. Georg Hass of-the U. S. Army Electronics Command, NightVision Laboratory, Ft. Belvoir, Va., and Mr. William Hunter of the Naval ResearchLaboratory, Washington, D.C., January1969. ..

1O1O

h c I lo9

(v Iz Y ri, Z

0& 0 E IO8 X' 3 LL

1 o4

1o6 1 o3

1o5 1 o2 0. ENERGY (keV) Figure 1 : TRAPPEDPROTON SPECTRA AT SYNCHRONOUSALTITUDE

49 Alternate Specimen Location

Turbomolecular

UV MONOCHROMATOR

J -1 ight Beam

IRRADIATION TEST CHAMBER

Mirror Test Specimen

Gas Discharge :ting Light Source xCold :ting Diffusion Pump Trap Figure 2 : SCHEMATIC OF PROTON RADIATION TEST FACILITY 1 .

Figure 3: REFLECTOMETER INTERNAL CONSTRUCT1 ON Light

"""I Reference P.M. Tube -" \ Monochromator Entrance SI it \\

""-

Tube (shown in incident and reflected Iight positions)

Figure 4: SCHEMATIC OF IN-SITUOPTICAL MEASUREMENT SYSTEM L I

Internal SignalInternal El ec tro- Photo- Photo Meter 6 @ Mu1tip! ier * Multiplier

Analog I

PowerSupply Voltage

High Voltage Analog (0-200 Millivolts) h i + Digital Vol tmeter

y I\ Plntter L From Potentiometer v I Circuit Tied to Location scattering photomul tipler Specimen Rotation of during scattering measurements Figure 5: REFLECTOMETERISCATTERING MEASUREMENTELECTRONICS I

11I -2 -1 Flux = 1.4 x 10 Protons - cm sec Before Irradiation 14 -2 ...... After 2 x 10 Protons - cm 40- 14 4 """ 5 x lol5 -.- 1 x 10 - 15 30 "- 2x 10 15 I "" - 5.3 x 10 5 " 16 x) 1 x 10

10-

I I I I I I 1 Ill1 IIII IIII 0 I I 90 100 20 300 4lM 500 &la

WAVELENGTH (NANOMETERS) Figure 6: REFLECTANCE OF A MgFZ/AI-COATED CER-VIT MIRROR IRRADIATED BY 10 KEV PROTONS I I I I I I I I IIII Ill1 1111 1'1 -2 -1 Flux 1.4 x 10 Protons - cmsec ' 1 14 -2 -I- ..I ...... " After 2 x 10 Protons cm 14 - """ 5x lol5 E -60 I- "- 1 x lol5 "- 2x lo 15 -.*. - 5.3 x 10 16 " 1 x 10

2ZU -10 "4LI

+2 I I I 1 1 I I I IIIItI I IIIIIIIII 90 100 200 301) 400 500 600 3 WAVELENGTH (NANOMETERS) Figure 7: REFLECTANCE CHANGES OF A MgF2/A I-COATED CER-VIT MIRROR IRRADIATED BY 10 KEV PROTONS I 1 I I I i n 0 o Fused Silica Mirrors 0 c X a Cer-Vit Mirron 11 -2 -1 Flux = 1.4 x 10 Protons - cm sec

-3a

-2c

-1 a

0 0 15 15 1 x 10 5x 10 1 x loT6

INTEGRATED FLUX (PROTONS - C"*) Figure 8: DIFFERENCE IN RADIATION DAMAGE OF MgF2/AI-COATED MI RRORS AS SHOWN BY REFLECTANCE CHANGES AT 210 NM I

90 "I 7- 80 /' - -

11 -2 -1 Flux = 1.4 x 10 Protons - cm sec -

Before Irradiation 14 -2 - **..-*.*-* After 2x 10 Protons -cm 14 I """_ 5 x lol5 -.- 1 x lol5 - I "- 2x lo 15 I -...- 5.3 x 10 - " 16 il 1 x10 10- -

0. I I I I I I I I I 1 I I I I1I1 I IIlllll 90 1010 200 300 400 500 600

WAVELENGTH (NANOMETERS) Figure 9: REFLECTANCE OF MgF /AI-COATED FUSED SILICA MIRROR IRRADIATED BY 10 KfV PROTONS 0 14 15 15 1 x 10 1 x 10 1.5 x 10

INTEGRATED FLUX (PROTONS - CM'2) Figure 10: RATE EFFECT IN IRRADIATION OFMgF21AI-COATED CER-VITMIRRORS AS SHOWN BY REFLECTANCE CHANGES AT 210 NM 1 I 11111 1 I I I I Irly I I I I I 11'1 I I I 1 I 1111 I

- -35 - 0 Cer-Vit Substrate 1at 109 nrnwavelength Fused at nrn wavelength B Silica } 111 r a"&. - A Cer-Vi, -30 at8210 nm wavelength 0 Fused Silica 0 -25 0 -2c

-15

-1 c

-5

Vacuum a """""" A -h n

+10 -

+15 I I IfIIII I I I I I Ill I I I I I1Ill I I I I I1111 I I 0.2 1 10 10 lo00 ACCUMULATED TIME AFTER IRRADIATION (HOURS) Figure 11: SUMMARY OF POST-IRRADIATION REFLECTANCE CHANGES FORMgF21AI COATED MIRRORS C) Cer-Vit Figure 12: ELECTRON PHOTOMICROGRAPHS OF UNCOATED, POL I SHED MIRRORSUBSTRATES . 60 90 -

80-

2w 70- 2 W % 60- 3 5 50- - Kanigen-Nickel Plated Beryllium Y m"1 "I Cer-Vi t w 40- act;: CY 30- 2U

I 1 1 I I I I I I 1 I I I I I I I I I I lllIllld 90 100 200 300 400 500 600 WAVELENGTH (NANOMETERS) Figure 13: REFLECTANCE COMPARISON OFMgF21AI-COATED BERYLLI UM AND CER-VIT MIRRORS a)Irradiated With 1 O1 Protons - At Flux of 11 10 Protons - sec” (10 keV)

b) Unirradiated Figure 14: ELECTRON PHOTOMICROGRAPHS OF IRRADIATEDAND UNlRRADlATEDMgF21AI-COATED FUSED SILICA MIRRORS 62 .

Figure 15:'TYPICAL SCATTERED LIGHT DATA FOR AN UNIRRADIATED MgF2IAl-COATED FUSED SILICA MIRROR AT500 NM WAVELENGTH Figure 16: TYPICAL SCATTERED LIGHT DATA FOR AN UNIRRADIATED MgFz/AI-COATED FUSED SILICA MIRROR AT253.7 NM WAVELENGTH r - SourceDecollimation: 3.70 Arc Min - DetectorDecollimation: 6.22 Arc Min - - - - W -

------

0 Fused Silica

@ Beryllium

-Jr Cer-Vit

, I I I I I Ill I I I I I IIf 1 .1 10 100 SCATTERING ANGLE FROM EDGE OF SPECULAR PEAK - (ARC MINUTES) Figure 17: COMPARISON OF SCATTERINGDATA AT 253.7 AND 500 NM WAVELENGTHS

65 /-”

FIUX = 1.3 x IO’ Protons - cm-2 sec-l Before irradiation 14 -2 ----- After 5 x 10 Protons - cm

7.- 1 x

”- 2 x IO~~ 15 “0.7 5x 10

” 1 x 10l6

n I I I I I I I I I I I Ill I 11.1 Ill1 m 300 400 500 600 WAVELENGTH (NANOMETERS)

Figure 18: REFLECTANCE OF A LiFIAI-COATED FUSED SILICA MIRROR IRRADIATED BY 10 KEV PROTONS 14 ----- After 5 x 10 Protons - cm

11 m Flux = 1.3 x 10 Protons - cm 5% 4 LU Ii\ 9-q"$\I. n

1 "++" " \/ I

1111 IIII 1111 +10 I I I 1 I I I t 1 Yo IO 201) 300 400 500 600 WAVELENGTH(NANOMETERS) Figure 19: REFLECTANCECHANGES OF A LiFlAI-COATED FUSED SILICA M I RROR IRRADIATED BY 10 KEV PROTONS I I I I I I I I - 11 -2 -1 Flux = 1.3 x 10 Protons - cm sec (10 keV)

15 15 16 0 1 x 10 5x 10 1 x IO

INTEGRATED FLUX (PROTONS - CM-3

Figure 20: GROWTH RATE IN REFLECTANCE DAMAGE AT SELECTED WAVELENGTHS FOR PROTON IRRADIATEDPROTONLiFIAI-COATED FOR MIRRORS . I I I I I I I I I 1I I I' "1""I- Integrated Flux = 1 x 10 15 Protons - cii2 (IO keV)- 10 -2 -1 1.2~10 Pmkms-crn sec 11 "- 1.3~10 12 . . , ...... 1,2 x 10

"""""""- I

I I I I I I I I I I I I I I I I Ill11 90 100 200 300 400 506

WAVELENGTH (NANOMETERS)

Figure 21: RATE EFFECT IN IRRADIATION .OF LiFlA I-COATED FUSED SILICA MIRRORS -50 I I I I I 11 Ill1 I I I I1 1111

- Flux = 1.3 x 10 11 Protons - cm -2 sec -1 (10 keV) - 16 -2 -- After 1 x 10 Protons - cm - “40- --.-----.After 14 Minutes in Air 8 c .. ..-----After Additimalin 38 Hours Air -

- - -30 .. -

w (3 \ - Z \.. - \.. -\.. .. - b -. -..

90 100 2010 300 40 5010 m

WAVELENGTH (NANOMETERS)

Figure 22: POST-I RRADl AT1 ON REFLECTANCE CHANGES OF LiFIAI-COATED MI RROR EXPOSED TO AIR . -50 I I I I 1 I I I I I I I ' I I II'I'II'' -

-40- Flux = 1.3 x 10 11 Protons - cm -2 sec-1 (10 keV)- 16 -2 -- After 1 x 10 Protons - cm -

-30 - .. .._....- After 62 Hours in Argon -

- -

-20 - -

L U -

tW - w2 a -

0 I I I I I I I I I I III 1111 II 90 IOO 201) 300 400 500 WAVELENGTH(NANOMETERS)

Figure 23: POST-I RRADIATION REFLECTANCE CHANGES OF LiFlAI-COATED MIRROR EXPOSED TO ARGON

. . t

I I I I I I I I I I

40 - Before irradiation -

0 1 I I I I I I 1 I I I I I I I I 90 110 130 150 1 70 190230 210 250 WAVELENGTH(NANOMETERS) Figure 25: REFLECTANCE OF A PLATJNUM COATED MIRROR IRRADIATED BY 10 KEV PROTONS +5 1 I I 1 I 1 1 I 1 I i 1 1 1 I 90 110 130 150 170 190 210 230 w) WAVELENGTH(NANOMETERS) Figure 26: REFlECTANCE CHANGES OF A PLATINUM COATED MIRROR IRRADIATED BY 10 KEV PROTONS lo5'

Pressure = 5 x 10-7 torr 1o4 t H2 Peak at - 7.5 x lo5 II

10;

" 80 10 20 30 40 50 60 70. ATOMIC MASS UNITS(AMU) Figure 27: RESIDUAL GAS SPECTRUM FOR REFLEC' 'OMEIER CHAMBER- WITH TRAP FI LLED (BASELINE SPECTRl MI

75 Pressure = .6 x H2 Peak at 5.3 x 10 Prr Spectrum Adjusted for Equivalence with Baseline Spectrum (See Figure 27) at 14 AMU 1o4

1 o3

1o2 I I II

10’

L 20 30 40 50 60 70 80 ATOMIC MASS UNITS (AMU) Figure 28: RESIDUAL GAS SPECTRUM FORREFLECTOMETER CHAMBER + - MONOCHROMATOR 76 Pressure = 3.6 x 10-7 torr H2 Peak at 4.3 x 106 Spectrum Adjusted for Equivalence with Baseline Spectrum (See Figure 27 ) at 14 AMU

I o4

1o2 I/

lo1

1 oo I 1 20 30 ATOMIC MASS UNITS (AMU) Figure 29: RESIDUAL GAS MASS SPECTRUM FOR REFLECTOMETER CHAMBER + ACCELERATOR BEAM, TUBE (BEAM OFF)

77 100

40 Flux = 1.4 x 10'Protons - cm-2 set" (10 keV) BeforeIrradiation 30 "" After 1 x lo1 Protons - cm-2sec-1 ---*- After Calcium Carbonate Cleaning 20

300 40'0 500 WAVELENGTH(NANOMETERS) Figure 30: EFFECT OF ABRASIVE CLEANING ON REFLECTANCE OF AN IRRADIATED MgF21AI-COATED MIRROR c L 1 00 I 1 I I I I I I 1 I I I I IIIIIIIIII

90 - -

“L.’.L.-=-” -

h t 0‘ - 6 & W a v - - Flux= 1.4 x 10 11 Protons - cme2 sec” (10 keV) -

Before Irradiation 16 - II I I After 1 x 10 Protons - cm-2 and 804 Hours in Air w -- a v) ..ll.l..l.lmm After Exposure to Fluorine for 33 Minutes - -

0 I I I I I I I I 1 I I I I IIIIIIIIII 90 10 200 300 400 . 500 600 WAVELENGTH (NANOMETERS)

Figure 31: REFLECTANCERECOVERY OF AM IRRADIATED MgF2/AI-COATED CER-VIT MiRROR EXPOSED TO FLUOR1 NE T- -70 x -60

u -50 6 Z 5 -40 2 .J LL 2 -30 Z w -20 80 Z a -10 U= 5 wo w2 n +io

+20

Figure 32: REFLECTANCE RECOVERY OF AN IRRADIATED NlgF2lAI-COATED CER-WIT MIRROR EXPOSED TO FLUORINE AS SHOWN BY CHANGE IN ARlR

a I 'Tc x tlce 100 aI I I I I I 1 I I I I I I 1 I I I IIIIIIIIIIII

Ref1ec tance

60 Initial Reflectance

-*-*- After Exposure to Fluorine for 33 Minutes 50

10 Change in Reflectance

WAVELENGTH(NANOMETERS) Figure 33: EFFECT OF FLUORINE EXPOSURE ON REFLECTANCE OF AN UNIRRADIATED MgFZ/AI-COATED MIRROR 9o

80 t /7 ,.4*) 0 0 0 70 - / 0 \ /' 60 - '\ " " ,' 0 50 - '\.-/

40 - Flux = 1.3 x 101' Protons - cm'*sec'l (10 keV)

Before irradiation 30 - ---- After 1 x 1016 Protons - cm-2 -----*- After Exposure to Atomic Oxygen for 5 Minutes 20 -

10 -

0; I I I I I I I I 11 I1I I I IIIIIIIII] 90 1010 200 300 400 5010 600 WAVELENGTH (NANOMETERS)

Figure 34: REFLECTANCE RECOVERY OF AN IRRADIATED MgF21AI-COATED CER-VIT . MIRROREXPOSEDATOMIC TO OXYGEN . Flux = 1.3 x 1011 Protons - cm-2 sec-l (10 keV)

Before Irradiation

""11 After 1 x lo1 Protons - cmm2

-=-=-=-=-=1 After Exposure to Atomic Oxygen for 5 Minutes -...-.....-...... After Exposure to Atomic Oxygen for 10 Minutes

WAVELENGTH(NANOMETERS) Figure 35: REFLECTANCE RECOVERY OF AN IRRADIATED LiF/Al - COATED FUSED SILICA MIRROR EXPOSED TO ATOMIC OXYGEN I I I I I I I I I I I I I I I I I I I I I IIIII~

"00 After 1OI6 Protons - cmm2 (10 keV)

0.0.0. After Exposure to Atomic Oxygen for 5 Minutes

D..." " After Exwsure to Atomic Oxygen for 10 Minutes t ,/-" /' \ \ '' f //

""""""""""""--- / i -i

I1 I I 1 1 I 1 I I I I I I I I IIIIIIIIIII~

901 1010 200 300 400 5010 600 WAVELENGTH (NANOMETERS) Figure 36: REFLECTANCE RECOVERY OF AN IRRADIATED LiFlAI-COATED FUSED SILICA MIRROR EXPOSED TO ATOMIC OXYGEN AS SHOWN BY CHANGE IN -AI? * v R . . . * I

-1 00 I I I I I I I I I I I I I IIIIIIIIIIIT

-90 n 0 10 2 x -80

WZle -70 W U Z a -60 -5 I-; Y -I WIJ. -50 (4? z_ 03 -40 VI G Z 4 -30 U z -20 2W n -10

0 ' 100 200 300 400 500 f WAVELENGTH (NANOMETERS)

Figu re..37:. CALCULATED EFFECT OF CONTAMINANT FILM THICKNESS ON CHANGE IN REFLECTANCE OF MgFdA I-COATED MIRROR (25 NM OF MgF2) - 0"- Experimental Data - 11 -2 -1 Flux = 1.4 x 10 protons-cmsec 16 Integrated Flux = 1 x 10 protons-cm'2 (10 keV) - -Calculated for 4.1 nm Contaminant Film Overlay - - - - -

4 """"""""-""-=~- """

I I 1 I 1 'I I I I I I I I IIII1lIlIIIL 3 100 2010 3010 4-00 500 600 WAVELENGTH (NANOMETERS) Figure 38: COMPARISON OF CALCULATED AND EXPERIMENTAL CHANGE IN REFLECTANCE FOR IRRADIATED, FUSED-SILICA MIRROR COATED WITH ALUMINUM AND 25 NM OF MgF2 V * . .

3; I I I I I I I I I I I I I I I I IIIIIIIIII 0- - - - Experimental Data 3- 12 -2 -1 Fluxs1,4 x 10 Protons -crn sec Integrated Fluxzl.5x Protons -cm2 ( 10 keV) I- -Calculated for 15 nm Comtaminant Film Overlay I-

I-

J I I I I I I 1 I I I I I IIIIIIIIiIII %? 200 300 400 500 600 WAVELENGTH (NANOMETERS) Figure 39: COMPARI SON OF CALCULATED AND EXPERIMENTAL CHANGE IN REFLECTANCE FOR IRRADIATED CER-VIT MIRROR COATED WITH ALUMINUM AND 25 NM OF MgF2 --a- -0- 9 Experimental Data

""""_ """"" """""""""""""

90 100 200 300 400 50'0 WAVELENGTH (NANOMETERS) Figure 40: COMPARISON OF CALCULATED AND EXPERIMENTAL CHANGE IN REFLECTANCE FOR IRRADIATED PLATE-GLASS MIRROR COATED WITH ALUMINUM AND 43 NM MgF2 . n I- 80-

ZW 2 70- W 11 Y W u 60 - Calculated From Optical Constants Z Qc + + + Measured (Cer-Vit mirror 50- 000 Measured (fused silica mirror -I L W cd 40 - CDA03% 3 11 30- ul ::I 0 90 100 200 300 400 500 600 WAVELENGTH (NANOMETERS) 100

90180 h5 2 70 W

Yn w 60 U Z Q t, 50 W Calculated From Optical Constants -I L - -0- -0. - Measured (plate glass mirror) 40 EL CD 0 5 3 30 LLI n m 20

a 9c

WAVELENGTH (NANOMETERS) Figure 42: CALCULATED AND MEASURED REFLECTANCE OF MgF$AI-COATED MI RRORS (43 NM OF MgF2) 8 1.

100

90 L

80 n I- Z g 70 W cb 60

-=0.333 + R=57.8

0 1 2 3 4 5 6 7 8 9 10 CONTAMINANT THICKNESS (NANOMETERS) Figure 43: REFLECTANCE OF MIRROR COATED WITH AI t 25 NM MgF2 t CONTAMINANT FIW WlTH COMPLEX REFRACT1 VE I NDEX (1.40 - ik) 0.7r I I1 I I I I 1.7 I I 1 I II I I I I I I I I I I I I

-1.6

C 6 -1.5 5 9- -Y I- -1.4" Z 3 (0 LL to 0 W i= Gi - I-x 0.3 - - 1.3 u

3 0.2 - E Ill I I1 I 1 1 I 1 I 1 I I I I I I I I I 8 100 150 20'0 250 300 c. (D -a 0 WAVELENGTH(NANOMETERS) I r Figure 44: ASSUMED OPTICAL CONSTANTS FOR CONTAMINANT FILM c) 1p 51: * w L Iu