Index Of Refraction Of Fused-quartz Glass For Ultraviolet, Visible, And .

Journal of Research of the National Bureau of StandardsVol. 53, No.3, September 1954Research Pape 2531Index of Refraction of Fused-Quartz Glass for Ultraviolet,Visible, and Infrared Wavelengths1William S. Rodney and Robert J. SpindlerThe ind ex of refraction of fu sed-quartz gJass was determined for 24 wavelength s fr om0.34669 to 3.5078 micron s, llsing the minimum-deviation m eth od . The whole ran ge \\"ascovered with a single instrum ent. The variation in index between samples was determ ined;no definite variations in di sp ersion were observed. Possible r elations be t wee n puri ty anclind ex ar e di sc ussed.1. IntroductionThe physical and optical properties of fusedquartz glass make its increased use in the opticalindustry highly desirable whenever it can be readilyproduced in optical quality. In composition it isdefinitely reproducible as compared with opticalglasses, it is extrem ely stable, and its coefficient ofthermal expansion is remarkably low. Its transmiSSlon limits, both in tIle ultraviolet and the infrared, are nearly the same as for the crystallinematerial and in addition it is fr ee from do uble refraction. Some specimens of this material, particularlyfrom European range finders, are of excellent opticalquality.The refractive index of fused-quartz O"lass hasbeeninvestigated in the ultraviolet and visible regionsby Gifford and Shenstone and by Trommsdorff ; alsosome measurements were made in the visible regionby Riedel, by Rinne, and by Watson. All of thesedata, together with some D-line indices by otherobservers, are referenced, compared , and anal Tsedby Sosman [1].2 For the infrared region some indices of fused quartz have been computed from refl ection data but experimentally determined dataseem limited to three-decimal-place values [2). Som epublished data [3) indicate different indices for specimens obtained from different sources, presumablymade by somewhat different m ethods. However,very little precise and self-consiste nt information haspreviously been obtained by use of identical procedures for different wavelength regions and fordi fl'eren t specimens.2. Samples InvestigatedSeven prisms of fused-quartz glass were used inthis investigation . One was made over 20 yearsago from fused-quartz glass produced by th e GeneralElectric Co. Its indices in t he visible region and itsdegree of homogeneity were carefully investigated [3),and it is used as a refractive-index standard for theprecise calibration of refractometers. A secondprism was made of domestic fused·-quartz glass of1 An acco unt of thi, work contai lting less ex perimental detail but giving theessential resnlts is pntlished in J. Opt. Soc. Alll. 44,677 (1954).2 Fignres in brackets indicate the li teraL ure references at the end of this paper.recent onglO , also produced by t he General E lectricCo. T'wo other prisms, A and B , were made fromthe product of the Heraeus Co. in German - andobtained by Stanley S. Ballard, who made themavailable for this investigation. A fifth prism wasmade by grindin g and pol ishin g windows on theperiph ery of a 6-in . disk recently manufactured bvthe Nieder Fused Quartz Co., Babso n Park, Mass.Ana ther prism was made by placing windows on theperipher T of a cylinder of H eraeus fused quartz(sample C), approximately 4 in . long and 2 in. indiameter. T he seventh sample was made availableby the Coming Glass vVorks; it is known to haveeXLremely good ul traviolet transmission characteristics but has a stronger absorption band around 2.6J1. than the other samples. All seven s pec imens areof near optical quality and free of noticeable striaeand bubbles, except the elisk, which has considerablestriae. Even ir. this case, howeve r, precise dat.a wereobtainable.3 . Instruments Used in Measurements'IVThe atts precision sp ectrometer was used for thevisible region and the Gaertner precision spectromcter for the ultraviolct and infrared. Thevisible-region measurements were also repeated onthe Gaertner sp ectrometer for calibra tion purpo es.D escriptions of these instruments, with photogr aphs,arc given in previous publications by the authors[4, 5). Figure 1 is a schematic diagram of thethe Gaertner spectrometer showing some ch anges.On both instruments the minimum-deviation method,with its desirable features of high aCCUl"acy andsimplicity of index computations, was used .The spectra used in these measurements were themercUl"Y and cadmium emissions for wavelengthsout to approximately 2.5 J1. and the absorption bandsof polystyrene beyond this region to approximately3.5 J1. .A lead-sulfide detector was used for the emissionline spectr a and a Golay pneumatic detector for wavelengths in the absorption spectra of polystyrene.T he incident flux was chopped at approximately 10cis to stabilize the zero, and the resulting signalwere amplified in both cases by a gated amplifiersupplied for the Golay detector. The amplifiedsignal was recorded on a recording potentiometer.185

ARM'lines of the visual spectrum as determined by theuse of the detector.The spectra are then scanned to identify otherwavelengths and to determine the approximate scaleposition of the minimally deviated beam for eachknown wavelength. Once an approximate positionis determined, hand settings are made by observing;the scale position for maximum or minimum deflection of the recorder pen. The scale readings areaveraged for greater precision.5 . DataFIGUHE1.Modified Gaertner infrared spectrometer .81 a nd Sz, entrance and exit slits, respectively; S, global' source; IVC conden singmirror ; I /, collimator mirror; 'l"', telescope mirror ; OJ chopper ; D , detector; A,a mplifier ; R , recorder. The prism P is rota ted at one-half the rota tion ra te of Tby use of t he gear s G I and G,.Index determinations were made on Heraeus.prisms A and B at or very near 24 and 31 C. Theaverage temperature of the room was controlledwithin 0.2 deg C and was determined frequentlyby means of a thermometer having its bulb near theprism. Temperature coefficients were then computed and small corrections were made to adjust toexact temperatures of 24 and 31 C. On the otherprisms, index determinations were made near 24 Conly.The refractive indices of the Heraeus B prism at24 C have been represented by means of the dispersion formula [6] ( - expressed in microns)n4. Experimental ProcedureThe index of refraction for visible wavelengths isfirst determined on the \Vatts instrument by th eusual minimum-deviation method. The errors forthese data do not exceed 1 X 10- 5 The deviationangles are observed and recorded. With the aid ofan auxiliary telescope the prisms are then leveledand centered with respect to the optical axes of theGaertner instrument. A second telescope is usedto set the prism, for a given line, usually the 0.6438j.L line of cadmium, at its minimum-deviation position. The table is now clamped in this position.The mirror acting as the telescope objective isbrought into the beam so that an image of this linefalls on the exit slit, causing a deflection of thepotentiometer pen. The scale position corresponding to maximum deflection of the pen is observedfor several pointings. The instrument is now setand clamped at the average of these readings. Asystem of gears, designed to maintain minimumdeviation by causing the prism table to rotate atone-half the rotation rate of the telescope anddetector, is now engaged and the prism table clampsr eleased.The design of the Gaertner spectrometer, asadapted for non visible radiation, does not allowdouble deviation , or even the position of the undeviated beam, to be measured. Consequently, it isnecessary to compute the average reading for th eundeviated beam. This is done by applying a fewdeviations, as observed visually on the Watts instrument, to the scale readings on the Gaertner forminimum-deviation settings of the corresponding2 ') 979864 0.008777808-"' -2- 0.01060984.0622496.00000 - -2(1 )and the goodness of fit of the formula is shown by thesmall residuals, B - c, as tabulated in table 1, andby their nearly random distribution . The predominance of negative residuals for the region 3 to 3.5/Lindicates that further improvement can be made.HO'wever, no least-squares solu tion was attemptedbecause at - 3.4188j.L, where the maximum negativeresidual was obtained, there is observational uncertainty because of the very broad nature of thisabsorption band , which tends to make the wave length of the minimum difficult to determine. Notethat a positive residual was obtained for the Corningprism at this wavelength (last. column, table 1) .The indices computed by formula (1) are, ingeneral, better than the observed values, and accordingly table 2 gives such indices for selected regularintervals of wavelength.In order to compare the indices for the variousprisms, th e indices as computed by formula (1)have been subtracted from all othors. Thesedifferences are listed in table 1 and plotted in figure 2.From the variation shown in this figuro, the differences in index of different samples seem clearlyestablished. It will bo noticed that the Heraeus,Nieder, and Corning samples agree in general inrefractivity with the data of Trommsdorff and ofGifford. Gifford's data woro taken on the fusedquartz glass made by Gifford and Shenstone, 'w ithparticular attempts at purity of material. It seemsthat the Heraeus B and Corning samples, maypossibly be freer of traces of other substances,because the common basic impurities would tend186