Spectrometric Analysis of Gamma Radiation from Fallout from Operation Redwing - Research and Development Technical Report Usnrdl-TR-146

Page 1

SPECTROMETRIC ANALYSIS OF GAMMA RADnTION l FROM FAL~ UT FROM OPERATION REDWING / Research and Development Technical Report USNRDL-TR- 146 SAN FRANCISCO 24* CALIFORNIA

Page 2

Gamma radiation emitted by selected fallout samples collected at Operation REDWING was spectrometrical.ly analyzed for the purpose of determining its characteristics (expressed in terms of absolute photon intensities for each disc ernible gamma line) as a function of time after shot. Since the absolute photon intensitiess are given and most -- samples represent known fractions of the area of the fallout ,1 collectors, one result is that dose rates in the vicinity of the collecting site can be calculated. =?’ Another important result of the analysis suggests that NaU becomes an increasingly important contributor to the radiation hazard from fallo-ut at early times after detonation

Page 3

This report is one of a series of technical reports on the gammaray spectra of fallout samples collected atvarious weapon tests. Previous data has been reported in this laboratoryts USNRDL-420, 31 August 1953, USNRDL- TR-32, 26 January 1955, and USNRDLTR-106, 27 August 1956. The sample study reported here was conducted in conjunction with Project 2.6.3, Operation REDWING. The work was done under Bureau of Ships Project No. NS 081-001, Technical Objective AW- 7, and is a continuation of that described * * as Subtask 6 in this laboratoryts report to the Bureau of Ships, DD Form 613, July 1956. Similar work is described under Program 9, Problem 1, in this laboratory’s Semi-annual Progress Report, 1 July to 31 December 1956, Progress Report USNRDL-P-1,l

Page 4

yjcy of the report were greatly appreciated: ~palytica! and Standards Branch of Chem@~ Technology Division The excellent liaison both in regard tq qprpples and informatio~ Members of the Electronics Sectioq ~e~p cooperative at all

Page 5

the Sprin~, 1956, weapons test series knovn as Cperation ments were made to fleterifiinethe spectrometric characteristics of the fallout radiationversus time and to compare these characteristics for consistency among samples from the same and flifferentshots. * The present work is a continuation of the spectrometric analyses of fallout @mna radiation which have been made at Operation UPS50T-IO!OTi:OiJat the Nevada Test Site in 195312 Operation CASTE at the Pacific proving Grounds in 1954&, and Operation TEAPCT at the Nevada Test Site in 1955 takln~ advantage of further refinements in instrumentation and analyzing technique. project 2.6.3 collected samples of various types at several ship, barGe and land stations and returne~ them to N5DL by air. This rapid return allowed analysis to begin as early as 51 hours after shot time--a notable impllshed at Operation CASTLE. Other samples requlrlng

Page 6

for each ener~y from the various samples, while in previous reports only relative line intensities were given. Since each sample constituted an accurately ~etermined portion of the material taken from a fallout collector of known area, an estimate can be made concernin~ the total gamma source strength per unit area in the vicinity Other than reportln~ quantitative information on photons emitte~ per square inch of fallout collection area, this experiment exeinplifie~better organization and preparation in other ways. The samples were premetry was never a problem. ..Thespectrometer was llwelltion throughout the course of the experiment. The instrument was improve5 by the substitution of a Moseley X-Y plotter for the itself was made use of a better Brown recorder. Finally the analysis simpler anfimore consistent by the way of measurin: peak areas.

Page 7

A gamma-ray photon is detectable only by means of the ener~y it loses in an interaction or interactions ~~~thmatter. Gamma-ray photons lose energy by three main processes: photoelectric effect, Compton effect, an~ each effect is dependent on the energy of the gamma-ray photon and the atomic number of tie detectins material. going any of these processes ultimately manifests itself in the molecular excitation of the 5etecting material. Yany organic crystals(such as anthracene, stllbene, an5 this excitation in the form of useful scintillations of light whose pre~omlnant color 5epen5s on the crystal. The intensity of the light flashes produced In these materials is dependent (in some cases, linearly) on the energy of the Incident photon (provide~ it is totally absorbed). These light pulses are detecteflby a photomultipller tube which converts them into electrical pulses.

Page 8

NaI(Tl) crystals are in wide use today as @n~~aray detectors because of their high efficiency and linearovercome by sealing them hermetically In thin-walled al~iFor this experiment a special sample hol~er was made for the-spectrometer which allowed the bottoms of the H’ sample containers to be placed 2 Inches above the top of “ the collimator (a l/2-inch hole through 6 Inches of lead) and centered without further adjustment. The q-inch in diameter by housing (Fig. 1) were the same sis of the data from Operation tube, preamplifier and as those used in the analyThe output of the preamplifier was fed into a Tracerwere analyze5 by an Atomic pulse height analyzer, the was provided by a helipot, The output pulses from the amplifier whose output pulses Instrument Company Model 510 variable base line bias o; which analyzer were fed Into

Page 10

from 10 countsmeter has 7 scales ran~in~ in sensitivity per secon5 full scale to 1000 counts per secon~ full X coordinate was varieclby the base line bias of the analyzer. A mollificationin the form of apotentiometer 1 in series with the X input allowe~ fine afijustments$0 be ma3e in matching pulse height to particular lines on the recor5ing paper. A zypical recording made recorder is shown in Ftc. 2. ‘i’hefallout samples were supplied curi.estotal activity. Descriptive information about sample, the effective fallout area it represents, an~ cloufisample consisted of material caug’nton a filter paper tl?rou3hwhich a representative sample of air was 5rawn as

Page 12

collect ana hold fallout material while reducing to a before and after the collecting perio~. linentriggered either by an increase in ra~ioactive back~round or by han~, the cover of the 5evice opens and a collectin~ tray rises into the wind stream for the preset exposure time (variable has an effective collectin~ area of about 2.5 square feet -dividedInto numerous cells to reduce loss of material du~ . to win~.. Sanples, other than cloud samples, were obtained from this type of collector mounted on converte~ liberty After collection the fallout material was flown back to the Laboratory an5 aliquote5 by Chemical Technology Division. The aliquots su?plisd to this section for analysis were placed in the polystyrene containers and prevented from shifting, In the case of granular or pulverized samples, by the addition of a small amount of paraffin. The machine was calibrated so that on “stan5ZLrd” gain the total absorption peak of a 100, 200, or SOO, etc., kev inch, etc., line on the recqrdlng paper. This calibration was accomplished first by careful zero-settin~ of the X-y

Page 13

recorder and analyzer, an5 ti-.enby varyin~ the Cain of the der until total absorption peal:sfrom staniiarclsof kno~?n priate positions on the recariiingpaper (Fig. 2). It was found that this calibration was very stable, requiring only an occasional small adjustment In the zero-set control of, the recor5er to allow for variation in paper placement and , rulinG defects on the paper. That the equipment was essentially stable in the Y direction was shown by perio~ic -“ checks with the 6@cycle test position of the count-rate - meter anilby the linearity of the decay curves on semiPulse height spectra were recorded for each sample on standard gain (3 ITEVfull scale) and on “highi’gain-- amplifier gain increase~ by a factor of four (0.75 Mev full scale)--in tnose cases where the strength of the EWnple HOW long the taking of periocllcrecordings of could continue after shot time before the count too low for statistically significant results course, on the original strength of the sample. On nearly all of’ different color ink, the standar~-~aln records, a calibrated was recorded on the same sheet In a while on the high-gain records a

Page 14

203 source was used.calibrated lIg Tnis use of the calibrate~ standards allowed a constant check to be made on the stabilIty of the equipment (see Appendix I). ANALYSIS OF DATA height spectrum of a monoenergetic gamma ray ‘ total absorption peak and continuous “Compton” If the energy of the gamma ray is above 1.02 Mev, pslr production begins to take place and one of two secondnry pepks Is contributed to, according to whether M or both of the annihilation quanta escape from the crystal. The peaks all have the shape of normal distribution functions or “Gaussians” because of the statistical nature of the photoelectric effect and the seconfiary-emisstonphenomenon inherent h the operation of a photomultlpller tube. If two Incident gamma energies are present, the lower is superimpose~ adiiitivelyon the continuous Compton distribution of the higher. An example of the idealized additton of three different monoenergetic gamma rays”is shown In Fig. 3. As more gamma energies are added the resulting confusion becomes progressively worse. (If the higher energies are above 1.02 Mev, the presence of escape peaks further complicates

Page 15

Fig. 3 Addltlon of Three Total Absorption Peaks at Different Energies and Their Associated Compton Distributions Showing the Resultant Spectm (heavy

Page 16

anfiarrive at the basic gamma lines, a method of graphical unfolding is use~. Gne starts at the highest-energy peak present (assuming that It is not superimpose~ on the Compton distribution of a still higher energy) an5 measures the area under the total absorpticm peak. Then knowtn~ the peak-tocribed later) for each energy, one can calculate the area of the Compton distribution. After drawing in the Compton distribution of the highest energy, the process Is repeated for the next highest-energy peak using the Compton of the * highest energy as a base line. The process iS continued until all the energies present are analyzed. the instrument for each energy, is only one energy contributing one can tell peak, ‘or two or more unresolveiienergies present, by wl~th of the peak. The resolution (width of a peak in ‘ at half-maximum height) was determined by measuring the resolution for several total absorption peaks of In or~er to facilitate the measuring of the areas and to aid in estimating the relative size unresolved peaks,a series of Gaussian curves of predetermine widths, various heights and,thus, known areas, were ~rawn by means of a Gaussian curve generator (see Appen31x II). BY

Page 18

superimposition of the suitable Gaussian on the pulse height spectrum with the aid of a froste3-glass light table, the area of each total absorption peak was determined without by this measuring with a planimeter. For errors lncurre3 system see the sec%lcm on errors. . absolute photon intensity to total absorpaome function of the energy of the gamma For the 4etimn$nat* of the graphical relationship between Z\A and E, ltuaa first necessary ascertain further machine parameters and other energyis simply the fraction of incident photons of a partlcular energy which Interacts In the crystal and Is Where ~ IB the total linear absoqptlon coefficient for gamma raya In soilhuniodide and X IS the crystal v versus E la shown plotted In Fig. 5. mined experimentally by-recording the spectra of several sources ti “knmn gamma energy and measuring areas of the trum. The-@e of’%hese areas platted agahst energy

Page 19

furnish a fairly smootlicurve. The curve characteristic of ’the machine experiment is shown in Fis. 6. (3T Counts per second per square inch total absorption peak, C. This number is a of the machine an~ Is foun~ by divitlingthe counts per secon~ per Inch of deflection In tion by the channel width In chart inches. width use5 throughout the experiment corresponds to 0.1 inch on the chart paper and the total Y 5eflection1s 10 inches. Then for the 200 c/s scale, C is 200. the fraction of photons leaving the source which enter the crystal an~ takes into account effects of collimator point source and are aiscussed in an opaque collimator. These effects some 5etall In a theoretical paper by 9 Derived from informa-R. L. Mlther of this laboratory. tion containe~ in this paperis Fig. 7, showing K, the source re5uction factor for a point source, versus r, the distance of the source from the collimator axis. To fin~ the average source re~uction factor, ~, to account

Page 20

Fig. 6 Ratio, w, of the Area Under the Total Absorption Peak to the Area Under the Entire Pulse-lieightDistribution Versus Gamma Energy

Page 22

to perform the extension of the source, it was necessary where r5 is the radius of the extended source. Tile function of r determined derive~ from Fig. 9 of equal to or larger than determination of ~ is somewhat simplified by that X Is constant over part of the ran~e of where a is the collimator diameter, b the colllmtor thickness And z the distance from the center of the ~ collimator to the source. Vhen r Is larger than a\2 - and smaller than az/b, K is some graphically by interpolation and Iflathertspaper. For values of r i.e.Jto include penetration effects, an approximation is used. Penetration is a function of Gamma-ray energy and the increased aperture due to increased energy is “approximatelythe same as the geometrical aperture of an opaque collimator two mean-free-paths less tnick. For the particular geometry involved in the present’ experiment, the values of the’machine parameters are as

Page 23

where l/v is the mean-free-path of gamma radiation tion for values of E (an~ corresponding values of i The absolute photon intensity Is a combination of these . A graph of I/A versus E is shown In Fig. 9. After rea~ing i{ the energetically appropriate I/A from the graph, the number was multiplle~ by the area of the total absorption peak of the correspon~in~ Gamma ray to obtain absolute photon intensity of that line. errors are listed photon intensities with their estimated for each gamma llne of each record in The abbreviate~ sample notation used for

Page 25

brevity in Tables 2 atid 3 is explained in ‘Mbla 1. Table recoriin;s antiT:ble 3 lists the photon intensities fourtl from the hi~h gain (4X) recor~fn~s. Because of the increase in resolvin; power some lines which were unanalyzable or un~etectable on the standard gain records became analyzable on the high Gain records. The data have been left in tabular form information more readily accessible to those In actual practice the analysis is more to make the who wish to use might be assu.nedfrom the description in the Analysis of Data section for the followinz reasons: (1) The shape of the Compton distribution is known only approximately for a Given energy even though Its area Is known fairly accurately. Even if the shape were accurately known, it would be mechanically cllfficult to draw such a distribution in with a predetermined area. For this reason a straight line is used for the Compton distribution, with the resultin& rectan~le having the proper area. It Is found, however, after drawin~ in several times and ~ettin~ the “feel” of the machine, the base line for each succeeding peak can be estimated with similar accura~y

Page 26

this approximation contributes to the errors becomlnS progressively greater (2) If two or more energies are together to be resolved, one can draw Gausstans of the proper wi~th beneath overall error, the toward the lower present too close two or more representing these unresolved energies, …