18 IRE TRANSACTIONS ON MILITARY ELECTRONICS January

TABLE II contain D/D, for the three models computed from (8)LIMITING FLUJX-DENSITIES AND DISTANTCE-RATIOS and the "Cygnus A" red-shifts of Table I. It is, of

course, possible to equate 1, with D. since the red-

NGC 1275 asl Cygnus A as Standard shifts of NGC 1275 and Cygnus A are relatively small.SX 1026 Standard D/D, Thus, the entries of Table II can be used to calculate

__________ 11_____ _ distances in light-years or parsecs as soon as some ac-

!lD/D / Model Model Model curacy in the distances of these two standard sourcesl/l~DD.(1()3)has been attained. M\/eanwhile, the entries of Table II

7 5.86 52.1 58.8 60.8 57.8 and (8) show that, if the red-shift of a Class II radio10 4.90 43.6 48.4 50.1 47.9

s20 3.46 30.8 33.8 34.4 334 source iS large and its luminosity-distance D isequated40 2.45 21.8 23.5 23.8 23.3 to the distance I computed on the static Euclidean uni-80 1.73 15.4 16.3 16.5 16.2160 1.22 10.9 11.4 11.5 11.4 verse hpothesis, then D will be underestimated if x>

_______ _______ nid over-estim ated if -1.

High-Altitude Measurements of X Rays and FarUltraviolet Radiation*HERBERT FRIEDMANt

Summary-Since 1946, the Naval Research Laboratory has weather may be largely attributed to differential heatingconducted basic research in the physics of the upper atmosphere of oceans and continents and seasonal changes due to theby means of high-altitude rockets. The program has emphasized inclination of the earth's axis rather than to any variabilityall areas of research, including atmospheric structure and com-position, the ionosphere, airglow and aurora, meteors, cosmic in solar output. There is considerable evidence, however,rays, and rocket astronomy. In the last area, which includes X that certain large-scale terrestrial weather patterns are cor-ray and ultraviolet radiation measurements, NRL scientists related with solar weather, perhaps through the medium ofhave contributed a major portion of the experimental informa- invisible radiation in the form of ultraviolet rays, X rays,tion available today. These comprise all the existing data on . . .solar X rays, the X ray and ultraviolet emissions of solar flares, or particles. The relationship between solar phenomenathe first spectrogram of the sun covering the ultraviolet region and the quality of radio communications is very direct.below 3000 A and subsequent extensions of the spectrum into The ionosphere waxes and wanes in direct response to thethe extreme ultraviolet, the first quantitative measurements of flood of solar ionizing radiation. Long-term observationssolar Lyman-a, and the discovery of ultraviolet nebulosities and of critical frequencies of radio reflections show a remark-the Lyman-a glow of the night sky. A recent success in photo- able correlation with sunspot number. It has now beengraphing the profile of Lyman-a with very high resolution opens established by rocket experiments that variations in X raythe way to the use of optical resonance absorption as a gauge ultaviolet e xperents associationsunspotsof atmospheric composition. This method may prove to be a and ultraviolet emissions are also associated with sunspotsmost powerful technique for analysis of the very high atmos- and the fundamental processes involved in production ofphere, well beyond the range of satellite drag measurements. the ionosphere are fairly well understood. From an astro-The purpose of this paper is to describe the experimental ap- physical standpoint, the rocket observations have greatlyproach used in accomplishing the radiation measurement pro- advanced our knowledge of the origins of the various types

of radiation in the solar atmosphere, and may ultimatelycontribute to our understanding of the sources of energy

INTRODUCTION available near the surface of the sun.BT) Y FAR the major portion of the effort that has gone

into rocket astronomy has been directed towar d ROCKET SPECTROSCOPYstudies of solar radiation and solar terrestrial rela- Spectroscopy became important to astronomy when

tlonships. Our weather, plant life, and natural energy photographic film was invented about 100 years ago. Sincesources such as coal, oil, wind, and water power are all de- then, almost all experimental techniques in astronomy haverived from the sun. The total flow of solar energy appears been based upon the use of photographic registration. It isto be very nearly constant, and the source of variability in not surprising, therefore, that the first instrument devel-

oped for rocket spectroscopy was a concave grating spec-* Mausciptrecive bythe GMJ, Otobr 2, 159.trograph in which the film, after exposure, was wound in-

t U.S. Naval Research Lab., Washington, D.C. side a light-tight cassette made of steel. The film had to be

1960 Friedman: High Altitude Measurements of X Rays and Far Ultraviolet Radiation 19

recovered after impact and subsequently developed. Afterearly successes had extended the known solar spectruminto the near ultraviolet,' it became clear that highly ac-curate pointing controls and photoelectric detection tech-niques would be necessary to measure the shorter wave-lengths of the ultraviolet and the X-ray end of the solarspectrum. Considerable success has been achieved in boththe instrumentation of sun-seeking devices and the devel-opment of electronic radiation detectors.

If a spectrograph is fixed to the body of a ballisticrocket, roll and yaw may reduce the effective exposuretime by a factor of 10 to 100 times. To compensate for theinherent instability of the rocket as a platform for a spec-trograph, the early instruments were designed with ex-tremely wide useful fields of view and succeeded in record-ing spectra down to 2100 A. The first pointing control wasa single axis sun follower compensating only for the rocketspin about its long axis. In recent years, a sun-followingbiaxial pointing control described by Stacy, Stith, Nidey,and Pietenpol2 has been used to compensate for yaw aswell as roll with considerable success. By means of a servosystem directed by a photoelectric angular sensing device,the spectrograph is kept pointing at the sun for a time ofthe order of 5 to 10 seconds with an accuracy of 1 minuteof arc. Over a period of 3 to 5 minutes, the error does not |exceed 5 to 10 minutes of arc. With this pointing device,the Lyman-a line of hydrogen at 1215.7 A was first Fig. 1-Rocket spectrograph mounted on biaxial sunfollower.photographed in 1952.8 Since then, the limit has beensteadily pushed to shorter wavelengths until it now appears By analogy, solar physicists refer to variable solar featuresfeasible to record even the X-ray spectrum. Fig. 1 is a as "solar weather." In spite of the great amount of detailphotograph of a spectrograph mounted on a biaxial point- observed in visible light, however, it is difficult to extrapo-ing control designed for the Aerobee rocket. At rocket late to a picture of the sun in invisible wavelengths. It istake-off, the entire spectrograph housing is protected by essential to be able to record images of the sun in the rangethe rocket's nose cone. At a height where air drag becomes of ionizing wavelengths including Lyman-a, the heliumsmall, protective panels are jettisoned and the biaxial resonance lines at 584 A and 304 A, and in various X-raypointing control swings the spectrograph out so that the wavelengths.photoelectric sensing eyes may seek the sun. Correction To photograph from a rocket requires even more precisefor the rocket's roll was accomplished by rotating the pointing than that developed for the spectrograph. Onlyentire nose section about the rocket axis. At the same time, with extremely fast optics is it possible to stop the residualcorrection for the yaw is achieved by permitting the spec- movement in the pointing device and to obtain sharp pic-trograph to swing in trunnions attached to the rotating tures. The photograph shown in Fig. 2 is the first detailednose section. image recorded in the ultraviolet at the wavelength of

Lyman-a (1215.7 A), the principle resonance line ofIMAGING THE SUN IN ULTRAVIOLET LIGHT atomic hydrogen. The Lyman-a camera is diagramed in

The face of the sun presents an ever-changing pattern Fig. 3. This instrument was the result of four years ofof visible events. Large numbers of phenomena are ob- development at NRL by Tousey, Purcell and Packer.' Theserved. Among the more familiar features are sunspots, only optical material which transmits the Lyman-a wave-plages, prominences, and flares. These phenomena are length is lithium fluoride. An earlier camera, designed byinterrelated in somewhat the same fashion that metero- Rense of the University of Colorado and utilizing prismlogical processes combine to produce terrestrial weather. optics, yielded a comparatively crude image but indicated

a much higher speed would be necessary to obtain clear'W. A. Baum, F. S. Johnson, J. J. Oberly, C. C. Rockwood, detail. This increased speed was achieved through the de-

C. V. Strain, and R. Tousey, "Soler ultraviolet spectrum to 88 velopment of very highly reflecting surfaces5 which per-kilometers," Phys. Rev., vol. 70, pp. 781-782; November, 1946.

'D. S. Stacey, G. A. Stith, R. A. Nidey, and W. A. Pietenpol,"Rocket-borne servo tracks the sun," Electronics, vol. 27, pp. 149- 'D. M. Packer, J. D. Purcell, and R. Tousey, "Lyman-alpha151; January, 1954. photographs of the sun," Nature, vol. 184, pp. 8-10; July, 1959.'W. A. Rense, "Intensity of Lyman-alpha line in the solar 'G. Hass and R. Tousey, "Reflecting coatings for the extreme

spectrum," Phys. Rev., vol. 91, pp. 299-302; July, 1953. ultraviolet," J. Opt. Soc. Amer., vol. 49, pp. 593-602; June, 1959.

20 IRE TRANSACTIONS ON MILITARY ELECTRONICS January

line image can be reproduced. Such techniques may makeit possible ultimately to transmit daily solar weather mapsin invisible ultraviolet and X rays from satellite observa-tories.

PEIOTOELECTRIC SENSORS FOR ULTRAVIOLET AND X RAYS

In parallel with its spectrographic program, the NRLdeveloped a series of narrow-band photoelectric detectorscapable of isolating important regions of the ultravioletand X-ray spectrum. The information from these sensorsis telemetered continuously during the flight of a rocketand traces the atmospheric absorption characteristic ofeach specific radiation band to the peak of the flight. Notonly do such experiments measure the solar flux, but theyidentify the region of the atmosphere which is affected. Ifthe absorption is characteristic of a particular constituent,the variation of intensity versus altitude is a measure ofthe particle density of that constituent. This technique hasbeen used to give the only experimental evidence of thevariation of molecular oxygen with altitude in the E andFl regions of the ionosphere. At X-ray wavelengths where

4 ~~~~~~~~~~absorption is essentially independent of composition, theradiation technique is a direct measure of atmospheric

Fig. 2-Photograph of the sun in the wavelength of density.Lyman-a (1215.7 A) March 13,1959. Experiments with narrow-band photodetectors were

begun by the NRL in 1949.6 In the decade between thenG"RATING and now, all the available information on solar X-rayfluxes below 100 A units has come out of the continuing

program of upper air research at NRL. In a similar way,the NRL program obtained the first quantitative measuresof the important Lyman-a emission and a major portionof all subsequent data on its variation over a decade ofobservations.7 In three major programs to study the emis-

-1SSt"6T66^DT'To' 1-. /, sions of solar flares, Project Rockoon (1956), Project Sun-flare I (1957), and Project Sunflare II (1959), narrow-

Fig. 3-Diagram of Lyman-a solar disk camera. band photodetectors have been used to measure the X-rayand ultraviolet emissions accompanying solar flares.8

mitted the use of mirror optics in the NRL instrument. Restriction of bandwidth in gaseous ionization chambers

ese were not ordinary mirrors, however but were and photon counters may be accomplished by selectingThesactionweretinot-pordinar mirrors,rhwever, t we combinations of ionizable gases and window materials with

diffratione gratihe inch mThe rulings caused the intense well defined transmission limits. Table 1 lists various win-15,000 lines to theinc the trlng cused the intens dows and gas fillings together with the ranges of spectralvisible light from the sun to be thrown out of the cameraleaving only the monochromatic Lyman-a radiation to response.form the solar image. The combination of high-speed At X-ray wavelengths, use is made of characteristic

ptics and the very intense emission of Lyman-ax radiation atomic absorption edges of thin foils and films to definefroptic sun made it possible to obtain satisfactoryfilm short wavelength cutoff. For example, 8 A marks the Kfrom the sun made It possible to obtain satisfactory film edeoalmnm nd2AthKegefcrb,wih

densities i exposure tmes of 1150of a second edge of aluminum and 24 A the K edge of carbon, whichdensities in exposure times of 1/50 of a second. i h ao osiun fMlradGytlpatcflsAt shorter wavelengths, reflectivities deteriorate rapidly IS the major constituent of Mylar and Glyptal plastic films.At horer avlenths rfletivtis dterorte apily On the long wavelength side, the transmission falls in-

and it appears necessary to replace photographic registra-tion with more sensitive electronic detection. Suitabledetection devices will be described below. It appears to be 'H. Friedman, S. W., Lichtman, and E. T. Byram, "Photon

entrely feasible to place a detector at the focus ofapara- counter measurements of solar X-rays and extreme ultravioletentirely feasblDe to place a detector at the fOCUS of a para- light," Phys. Rev., vol. 83, pp. 1025-1030; September, 1951.bolic mirror telescope and to wobble the mirror so as to 7E. T. Byram, T. A. Chubb, H. Friedman, J. E. Kupperian, Jr.

and R. W. Kreplin, "Intensity of solar Lyman-alpha and adjacentscan the solar image across the detector. If the wobbling ultraviolet emission lines," Astrophys. J., vol. 128, pp. 738-741;movement is in the form of a TV raster, the response of November, 1958.

'H. Friedman, "Rocket observations of the ionosphere," PROC.the detector can be telemetered to ground where a multl- IRE, vol.47,pp. 272-279; February, 1959.

1960 Friedman: High Altitude Measurements of X Rays and Far Ultraviolet Radiation 21

TABLE I solar radiation above the atmosphere, this tube respondsNARROW-BAND PHOTON COUNTERS AND strongly to the solar flux of about 3 X 1010 quanta cm-2 S',IONIZATION CHAMBERS between 1425 and 1500 A, while rejecting the photoelectric

Window Thickness Ionizable Response Band contribution of about 5 X 1012 quanta cm-2 s-1, betweenGas (Angstroms) 1500 and 2000 A. This spectral response characteristic isBeryllium 0.005 inch neon 1- 8 ideal for measuring the absorption of solar radiation byAluminum 0.00025 neon 8- 18 molecular oxygen in the dissociation continuum.10Mylar 0.00025 helium 44- 60Glyptal 0.00006 helium 44- 100 Between 100 A and about 1100 A only the very thinnestLithium fluoridie 2 mm nitric oxide 1100-1340 films transmit appreciably and it is difficult to prepareCalcium fluoridie 2 mm nitric oxide 1225-1340Sapphire 1 mi xylene 1425-1500 windows that can meet the requirements of vacuum tight-

ness and mechanical strength. A nitrocellulose film aboutone thousand angstrom units thick transmits about 44 per

versely as the cube of the wavelength. In an X-ray photon cent at 46 A and about 17 per cent at 220 A. An evaporatedcounter or ion chamber, the ionization current is derived film of pure aluminum 500 A thick begins to transmit atfrom photoelectric absorption in the gas. To avoid response 830 A on the long wavelength side and remains fairlyto wavelengths very much shorter than the K edge of the transparent down to the L III edge at 170 A. Purcell andfilter, where transmission again becomes appreciable, the Tousey have suggested using such a film over a layer ofgas is choseni from one of the lighter gases such as helium flourescent material coated on the window of a photomul-or neon, which are transparent to the shorter wavelengths tiplier tube to provide detection of wavelengths within thebut absorb strongly at wavelengths longer than the K edge transmission range of the aluminum film. It is also possibleof the filter. to dispense completely with the window material and to

Gaseous ionization detectors operating in the ultraviolet use photon counters or ionization chambers of the free-region above 1000 A exhibit a spectral response com- flow type. The solar spectrum is sufficiently intense thatpounded of: 1) a long wavelength surface photoelectric measureable responses could be obtained through a windoweffect with a threshold above 2000 A for most metals; 2) as small as 0.005 inch in diameter. A small flask of gas mayan internal photoelectric effect confined to wavelengths be used as a reservoir to maintain constant pressure withinbelow 1500 A; and 3) photoionization of a gas. The yield the photon counter as the gas flows out through the windowof the long wavelength surface photoelectric effect is orifice. By selecting the fill gas from the rare gases, it issmall, of the order of 10-s to 10-7 electrons per quantum. possible to control the long wavelength threshold in aFor most metals a new threshold appears at a wavelength number of steps. The ionization potentials of helium,in the far ultraviolet, usually between 1000 and 1400 A neon, argon, krypton, and xenon are 507, 577, 791, 890,at which the yield may multiply abruptly by as much as and 1027 A respectively. A helium flow counter at low1000 times. 'Finally, most gaseous molecules have thresholds pressure makes an excellent monitor of the He II reso-for photoionization below 1500 A and reach such high nance line at 304 A.cross sections for photoionization that yields of close to The simple vacuum photocell would be an effective de-100 per cent are possible.9 tector for the 100 to 1100 angstrom range if it were possibleA wide variety of filter materials and gas absorbers are to suppress the long wavelength response. T. A. Chubb of

available for the construction of narrow-band photon NRL has succeeded in preparing tubes with evaporatedcounters. Tatbles II and III list the transmission properties lithium fluoride surfaces that exhibit yields of 40 per centof certain solids and gases. How to combine the various at 585 A and less than 1 per cent at Lyman-X. Hintereggerproperties to produce a narrow-band photon counter may of the Air Force Cambridge Research Center has appliedbe illustrated by specifying the selection of window and a method of retarding potentials to scan the photoelectrongas filling for a tube sensitive to wavelengths near 1450 A. energy distributions between Lyman-a (1216 A) and HeThe window may be synthetic sapphire which is transparent II (304 A). If it is desired to use secondary emission mul-down to a short wavelength limit of 1425 A. Xylene vapor tiplication after the photo-surface, the dynodes must beis included because it has a photoionization threshold at prepared of materials which do not respond to longer wave-1500 A. Nitric oxide is nonphotosensitive above 1350 A lengths. A silver magnesium surface is insensitive to lightbut is added to the gas mixture because it effectively of wavelength longer than 3000 A and still provides sec-captures all electrons released by long wavelength photo- ondary emission multiplication factors in excess of 2. Aelectric effect on the cathode surface. The combination of multiplication surface consisting of a tin oxide layer ona sapphire window and a gas mixture consisting of 44 mm glass has recently been developed by the Bendix Researchof nitric oxide, 0.5 mm of xylene, and 650 mm of helium Laboratories which exhibits almost no photoelectric re-produces a spectral response which is confined to the sponse at wavelengths longer than Lyman-oc.region between 1425 A and 1500 A. If used to measure Photon counters can be produced with sensitivities of

10E. T. Byram, T. A. Chubb, and H. Friedmnan, "Photon count-9T. A. Chubb and H. Friedman, "Dissociation of Oxygen in the ers for the far ultraviolet," Phys. Revo., vol. 98, pp. 1594-1597;

upper atmosphere," Rev. Sci. Instr., vol. 26, pp. 493-495; May, 1955. June, 1955.

22 IRE TRANSACTIONS ON MILITARY ELECTRONICS January

TABLE IITRANSMISSION CHARACTERISTICS OF SOLID MATERIAL IN THE VACUUM ULTRAVIOLET

Approximate Wavelength Regions Wavelength RegionsApproximate in Which Trans- of Less than 1 PerMaterial Filter mission Exceeds Cent Trans- Character of Short-wave CutoffThickness 10 Per Cent mission

mm (Angstroms) (Angstroms)

LiF* 0.4 > 1050 <1030 Sharp with moderate absorption up to 1300 A

X-rayed LiF 1 1200-2200, >2800 2300-2700, <1175 Depends on F center density

X-rayed and thermally bleached LiF 1 1100-2000, >2400 2100-2300, <1075 Similar to uncolored LiF

CaF* 3 >1220 <1215 Sharp

Evaporated film CaF2 on LiF base > 1200 < 1150 Good transmission at X > 1200 A compared tothat for X1150-X1200 A

Sapphire (synthetic)t 0.5 >1425 <1415 Sharp

Fused quartzt 1 >1560 <1525 Gradual. General reduction of UV transmis-sion by X-ray irradiation

Topaz (natural) 1 >1550 <1525 Sharper than fused quartz

Gypsum (natural) 1 >1620 <1600 Similar to topaz

NaCl* 1 >1710 <1700 Sharp

KCI* 1 > 1750 < 1740 Sharp

KBr* 1 >2010 <2000 Sharp

Teflon 0.006 > 1700 < 1660 Gradual-considerable loss of light due to scat-tering

SrF2|j 0.88 >1300 < 1270 Sharp

NaF* 1.37 >1310 <1276 Sharp

* Harshaw Chemical Co.t Linde Air Products.t Central Scientific Co.Optovac Co.

TABLE IIITRANSMISSION CHARACTERISTICS OF GAS FILTERS FOR

THE VACUUM ULTRAVIOLET

Approxi- Approxi-athicknltes Regions of useful Regions of high mate filter Regions of useful Regions of high

Gas i ,cm(es transmission opacity Gas thickness transmission opacityIn cm (re- (Angstroms) (Angstroms) |in cm (re- (Angstroms) (Angstroms)duced to (nsrm)(ntos)duced toNTP) NTP)

02 6.0 1102-1110, and nar- Other wavelengths SO2 0.1 1625-1750, >2250 1800-2150, <1590row transmission shorter than 1750bands centered at H2S 0.1 1600-1700, >2300 1800-2200, <15751124, 1158, 1166,1189, 1216; >1800 CO2 0.5 1175-1250, >1650 1300-1550, <1150

CH3C1 0.1 1425-1460, >1850 1475-1610, <1420 CCl4 0.0025 1160-1200, >1550 1220-1330, <1150

CH3Br 0.05 1525-1575, >1800 1610-1775, <1520 C12 0.1 About 12 transmis- 1340-1420, <1170sion bands between

CCI2F2 0.0025 1200-1230, >1375 1234-1325, <1195 l170Oand 1310. One~~~~~~~~~~~~~~~bandtransmits Ly-

CS2 0.05 About 10 transmis- 1800-2100, <1520 man radiation;sion bands between <14501530 and 1780,>2200 CH4 0.025 >1400 <1375

NH2 0.1 About 10 transmis- 1700-2050, <1425 C3H8 0.1 >1600 <1575sion bands between -. __ _________1450-1700; >2150 (CH2)3CH 0.1 >1060 <1640

N20 l0.025 1200-1220, >1530 |1220-1350, <1190 C2H4 |0.025 |>1850 |<1750

1960 Friedman: High Altitude Measurements of X Rays and Far Ultraviolet Radiation 23

10-' erg cm-2 s-' at a signal-to-noise ratio of 10. Such de- likely that this spectral line will provide a good measure oftectors have rnade it possible to initiate the science of rocket molecular nitrogen at heights above 200 km.astronomy with unguided ballistic rockets.1" Fig. 4 is a con- Although the above examples are impressive evidencetour map of a portion of the sky in the neighborhood of of the power of the optical approach, a high-resolutionOrion mapped by a photon counter sensitive to the band spectroscopic technique developed by Purcell and Tousey1225 to 1350 A. The sensitivity of the detector can be ap- at NRL points the way to an extension of the sensitivitypreciated when it is pointed out that the flight time above and selectivity of optical absorption measurement by sev-the absorbing terrestrial atmosphere was less than 2 min- eral orders of magnitude. The cross sections involved inutes and that the tube was fixed to the skin of a rocket that the above-mentioned examples fall in the range of 1J20 towas spinning and precessing freely without constraints. 10-17 cm2 and are associated with processes of ionization

and dissociation. In resonance absorption, however, thecross section may be greater than 10'13 cm2 in the center

7-20' _ of the absorption line. In principle, it should be possibleO (< lto determine the concentrations of hydrogen and various

U(-2 27Ac>. < other atmospheric constituents to heights of several thou-+10, f0 sands of kilometers by measuring the amount of self re-

versal in the core of the appropriate solar emission line asa function of altitude of a vertical rocket probe.

At the temperature that may exist in the very highatmosphere, a resonance absorption line should have a

-I0tQ4D1<t< , }width of the order of 0.1 A or less. Obviously, the measure-ment of the profile of such a line requires more than anorder of magnitude improvement in resolving power over

/,L *the spectrographic techniques that have previously been72oe.' 6h 5h 4h applied in rockets. Purcell and Tousey approached the prob-

Fig, 4-Ultraviolet nebulosity in the region of Orion. Isophotes of lem by using a 12,000 line/mm grating of 50-cm radius in1300-A radiation; valucs of surface brightness in units of 10 the thirteenth order of diffraction, since resolving power isergs/cm2.

directly proportional to order of diffraction. The theoreticalresolving power was 266,000, corresponding to approxi-

ATMOSPHERIC STRUCTURE mately 0.005 A at Lyman-a, and the dispersion of the in-Second onily in fundamental importance to mapping the strument was 2.6 mm/A. In the flight experiment, the

spectrum of the sun and other celestial sources is the ob- spectrograph appeared to resolve 0.03 A and showed a deepservation of the absorption of such radiation in the ter- narrow core of the order of 0.05-A width in the center ofrestrial atmosphere. The significance of the various ultra- the solar emission line about 0.7 A wide. From this profile,violet and X-ray wavelengths for the production of the it was deduced that the total neutral hydrogen content perionosphere has already been mentioned above and is treated square centimeter column between the rocket at 200 kmextensively in the literature.12 Using the sun as a light and the sun was between 1012 and 101 .source and mneasuring the variation in atmosphere attenua- Similar measurements may be made on atomic oxygention with altitude, it is possible to determine the height dis- and atomic nitrogen resonance lines. If the photographictribution of all the major and some minor constituents of film is replaced by a photoelectric sensor, the profile may bethe atmosphere to its outermost limits. One of the earliest scanned at various altitudes and telemetered so that it issuccesses of rocket spectroscopy was the detection of the possible to determine the atomic concentration vs altitude.ozone distribution up to 70 km. The attenuation of Ly- Such instrumentation is now being developed for flights inman-e is a gauge of the 02 concentration in D region and small rockets such as the Javelin and Journeyman, capablemay also provide a means of determining H20 at those of reaching altitudes of 1000 to 2000 miles.altitudes. Molecular oxygen has been traced to 170 km Although using a high order of diffraction to obtainby observing the absorption of solar 1500-A radiation and resolution has given encouraging results, an instrumentthe total atmospheric density has been measured to 160 capable of much higher resolution is the echelle spectro-km by using X rays in the 44- to 60-A band. Spectra taken graph. Such a grating has a higher blaze efficiency, greaterat a height of 200 km show no evidence of Lyman-v, which light gathering power and can be obtained with extremelyappears to be absorbed by molecular nitrogen. It seems low ghost intensities.

11T. A. Chubb, E. T. Byram, H. Friedman, and J. E. Kupperian, RAITO MNTRNG OCESNDAELTSJr., "The Threshold of Space," Pergamon Press, London, Eng., Rtrigt osdrto ftesml arwp. 203; 1957. Rtrigt osdrto ftesml arw

12R. J. Havens, H. Friedman, and E. 0. Hulbert, "sThe Jono- band photometers, it is clear thatmuch valuable informationspheric F2 Region," The Physics of the Ionosphere, Rept. of 1954Cambridge Coynference, p. 237. on solar variability and atmospheric structure can be ob-

24 IRE TRANSACTIONS ON MILITARY ELECTRONICS January

tained by their use. In the author's opinion, observations solar batteries distributed around the periphery of the spin-of the attenuation of solar radiation in the atmosphere con- ning satellite. The total weight of the satellite need not ex-stitute one of the most direct sources of information on ceed 200 pounds, and a satisfactory orbit would require aatmospheric structure. It is desirable to devise basic photom- perigee greater than 350 miles and an apogee less than 600eter packages for small rockets that can be flown in a lati- miles. The spin frequency should be about 0.5 rps.tude survey from the Arctic to the Antarctic and used in Consider the scientific data that could be provided by asynoptic studies to observe diurnal and seasonal variations. radiation monitoring satellite.

Instrumenting a satellite with a variety of narrow-bandphotometers covering as many significant wavelength re- Solar Emissionsgions as possible is also highly desirable. Observations The continuous recording of X-ray and ultraviolet radi-could be made of the sun, the night airglow, day airglow, ations would permit direct correlations of ionospheric be-and auroral light. havior with the effective solar radiations and relate the in-

Consider the possible makeup of such a radiation moni- visible radiations to visible phenomena such as flares, surgetoring satellite as conceived by the author's colleagues at prominences, and levels of plage activity. Relationships be-NRL. The solar photometers would consist primarily of tween solar radiation and the airglow and aurora wouldion chambers and photocells. Most of the X-ray detectors also be revealed.would be vacuum tight ion chambers fitted with thin win-dows and utilizing inert gas or nitrogen fillings. In the Attenuation of Solar Radiationspectral bands where no suitably transparent windows are Measurements of the attenuation of various solar bandsavailable, for example between 100 and 1050 A, low- during passage of the satellite into or out of the earth'spressure free-flow ion chambers could be used. These ion shadow should provide information about the vertical dis-chambers would be supplied with vapor obtained from tribution of atmospheric constituents such as 03, 02 and 0.reservoirs carried in the satellite. A free-flow ion chamber With a polar orbit, these data would be obtained initially atwould probably be used in the 1425- to 1500-A band where latitudes near the north and south poles and at all longi-photochemical decomposition of the ionizable gas is partic- tudes. As the earth moves around the sun in its orbit, a greatularly severe. The longer wavelengths, for example those variety of attenuation geometries would become possible atbetween 1600 and 5000 A, would be monitored by photo- all latitudes. The data would be limited somewhat by theemissive cells containing surfaces of CsI, Rb2Te, and Cs3Sb. fact that the sun has an angular diameter of half a degree.In the infrared, a photoconductive cell may be used. Aspect However, the resolving power in altitude would permit sig-information would be obtained on the basis of signals nificant world-wide comparisons of vertical distributions.furnished by an optical sensor which measures the anglebetween the satellite spin axis and the sun. The addition Day irgiowof the magnetic aspect system would provide a measure of The earth should appear to be black in the wavelengththe angle between the satellite spin axis and local magnetic region of ozone absorption 2400 to 2700 A except for radia-field. The combination of solar aspect and magnetic aspect tion by 02. It is of interest to see if this is really true.

permits a unique determination of the satellite aspect at Strong airglow emissions expected in the daytime are 3914any instant of time. A from N2+, 7619 from O2, 6300 from 01 and 5893 fromThe general method of making measurements would be Na. All of these are expected to show strong enhancement

based on utilization of satellite spin as a means of modu- near the horizon. The heights of these layers could belating the incident radiation fluxes. The satellite could be mapped on a world-wide basis.designed with a major moment of inertia corresponding tothe direction of spin, that is, in the form of a flat cylinder Zodiacal Lightor octagon with the mass concentrated near the rim. All the A detector, which measures a broad band of visible lightsolar detectors would be mounted looking out perpendicu- excluding the brightest airglow emissions, can provide datalar to the axis of spin. All night-time and day airglow de- on the zodiacal light in the ecliptic plane. The earth itselftectors would be paired at diametrically opposite positions would provide the necessary eclipsing disk before the sun.and directed at 450 to the spin axis. A nearly polar orbitwould permit the most interesting variety of scanning ge- Night Airglowometries for both the observation of atmospheric structure There are a number of characteristics of the night air-and determination of the movements of the airglow. glow that could be determined with photoelectric photom-

All of the required photometry could be accomplished by eters. First, the waves and variations that are known toa collection of about 50 sensors. The tracking transmitter exist in the airglow could be mapped. They are thought towould be similar to the Vanguard model radiating 10 mw be produced by waves in the circulation of the upper atmos-at 108 mc. Telemetry requirements would be satisfied by a phere, and such a map would give important data bearingpower amplifier radiating 4 watts on command and modu- on the circulation pattern. Each photometer would scan alated by 5 subcarrier oscillators. The required power for curved belt, broadest near the equator and narrowing some-all circuit functions would be obtained from 430 cm2 of what at higher latitudes. Each orbit would cover successive

1960 Arditi: A Gas Cell "Atomic Clock" as a High-Stability Frequency Standard 25

strips of the earth about 30° apart in longitude. Since the Ha line and for X rays in the 10 to 150 kev range. Theearth would be scanned always at the same local time, the world-wide distribution for Ha, Lyman-a, X rays, and thedata would be free from diurnal effects. A second result usual auroral emissions could be correlated.would be the height of the various airglow layers, from a X-ray and Lyman-a photometer packages have been pre-measurement of the zenith angle of the maximum in lumi- pared by NRL for several Vanguard attempts and for thenosity. A third result would be the study of twilight en- ABMA International Geophysical Year heavy payload.hancement for the various emissions, when the satellite Only Vanguard III succeeded in orbiting and it carries justpassed into twilight at high latitudes, both north and south. one channel of X-ray information. These past a'tempts rep-The measurements would cover the permanent artic twi- resent only minimal experiments and it is to be hopedllight regions as well as the transitory diurnal twilight. that full-fledged photometry satellites of the type sketched

above may soon be included in the national space effort.Aurora The experimental background for design of the photo-The photomnetry satellite offers the possibility of directly metric sensors derives directly from past rocket experi-

comparing the Aurora Borealis and Aurora Australis. The ments. In applying these techniques to satellite instrumen-distribution could be plotted relative to the earth's mag- tation, much more severe specifications are imposed on thenetic poles. C.orrelations could be sought between observa- stability and life of the sensors, but no unsurmountabletions of solar variations and changes in the auroras. These difficulties seem to bar the way toward satisfying all themight be expected to be most conspicuous for the auroral requirements.

A Gas Cell "Atomic Clock" as a High-Stability Frequency Standard*MAURICE ARDITIt

Q UARTZ crystal oscillators have been developed toH E

a high degree of accuracy. Stability of 3 X 10-9 | D1 T

per day or better has been achieved. These oscil- FREQUENCYlators, however, require an aging period of several months. ( 9 MUERThis basic defect can be greatly alleviated by locking thecrystal oscillator to the center frequency of a stable atomictransition through a servo-mechanism system. This is the C

general principle which is used in the atomic beam fre-quency standlard and also in the present gas cell standard. SERVOMoreover, a gas cell type of "atomic clock" is quite easily COfRe sadaptable to a small and simple package, and this is ad- Fig. 1-Simplegas cell atomic frequency standard.vantageous for airborne operation. good short-term stability (crystal oscillator and multiplierFollowing the work pioneered by R. H. Dicke and T. R. chain), a phase modulator, a gas cell where the atomicCarver of Princeton University, a gas cell frequency stand- transition takes place, a phase sensitive amplifier-detectorard has been developed by ITT Laboratories and recently and a feedback-loop servo locking the crystal oscillator totested for accuracy and stability.' Some of the results ob- the resonant frequency of the atomic transition. The mi-tained in these tests are reported here. crowave energy is frequency modulated at a low rate and

with a small frequency excursion, thus obtaining the de-G;ENERAL DESCRIPTION OF A GAS CELL rivative of the resonance line at the output of the phase de-

FREQUENCY STANDARD tector. In order to lock the oscillator to the atomic reso-A siplegaseel frquecy tandrd 1] illincudethenance, the signal output of the phase detector is fed back,Aollsimpegacmoellsfreqenc st.):andr [1]cillaoincud ther in proper phase, to an element controlling the freqluencyfollwin comonets see ig.1): n ocilatorof ery of the crystal oscillator, through an amplifier and servo-

control system.* Manuscript received by the PGMIL, October 29, 1959. For an atomic frequency standard of high stability andt ITT Labs., Nutley, N.J. high accuracy, the following three characteristics are es-Wokwas supported in part by the ONR under Contract- No. seta.

Nonr-2553(OO) NR 374-901. snll