y 'S9a 30iun»48 MDDC - 890 UNITED STATES ATOMIC ENERGY. COMMISSION NEUTRON MONITORING BY MEANS OF "SPECIAL FINE-GRAIN ALPHA-EMULSION" FILM by J. S. Cheka This document consists of 8 pages. Date Declassified: April 3, 1947 This document is for official use. Its issuance does not constitute authority for declassification of classified copies of the same or similar content and title and by the same author(s). Technical Information Division, Oak Ridge Directed Operations Oak Ridge, Tennessee 12-227-Coro r Digitized by the Internet Archive in 2011 with funding from University of Florida, George A. Smathers Libraries with support from LYRASIS and the Sloan Foundation http://www.archive.org/details/neutronmonitorinOOusat NEUTRON MONITORING BY MEANS OF "SPECIAL FINE-GRAIN ALPHA-EMULSION" FILM By J. S. Cheka Fine-grain alpha- emulsion film is used at Clinton Laboratories to monitor personnel for neutron exposure. Part of the film is behind a cadmium shield, which removes thermal neutrons; part of it is exposed without a shield, intercepting both thermal and fast neutrons. Calibrations were made to determine the sensitivity of the emulsion for both thermal and fast neutrons in terms of proton tracks observable under a microscope of 970X magnification. The ratio 3.01 x 10 a n^}/ observed track was noted for the N 14 (n,p) C 14 reaction, and the ratio 2.05 x 10 4 nf/ ob- served track was noted for proton recoils. Evaluating expected track densities, based on N-atom density for nth and H-atom density for nj, the first process was found to be about 53% effective and the latter 46% effective in producing recognizable tracks of 3 grains or more in this emulsion. Using tolerance values of 4700 nth/(cm 2 sec) and 266 nf/(cm 2 sec),* two-weeks tolerance doses are indicated by 39 tracks/ 50 fields for n^, and 33 tracks/ 50 fields for nj. However, since different batches of the same type of film were found to differ in sensitivity, each batch must be calibrated to establish the track density which indicates a tolerance dose. INTRODUCTION Alpha-particle-sensitive film has been used at Clinton Laboratories since October 1944, to mon- itor personnel for neutron exposure. This film is now in the regular film badge, in addition to the beta- and gamma-ray-sensitive film. This practice is possible because this film is sensitive to protons. Fast neutrons produce recoil protons in the hydrogenous emulsion of the film. Thermal neutrons produce protons in reaction with nitrogen, N 14 (n,p) C 14 . Part of the film in the badge is shielded with cadmium and part of it remains unshielded. Thus it is possible to use one film for both fast and thermal neutron monitoring. The present series of experiments included qualitative tests to determine the relative effective- ness of thermal and fast neutrons in producing protons sufficiently energetic to generate recognizable tracks in the emulsion. Distribution of track lengths in terms of number of silver grains per track was also studied. Quantitative tests were made, using various fluxes of both fast and thermal neu- trons. EXPOSURES AND TECHNIQUES The source of the neutrons was the Clinton pile, which is a graphite pile with a uranium lattice. The exposures were made in a tunnel on the top of the pile. This tunnel has thirty inches of graphite between it and the nearest uranium. It is also shielded by four inches of lead. The graphite serves *nf taken at 1 Mev. MDDC - 890 [1 -»»7-pl-»f-e-bu 21 MDDC - 890 as a moderator and slows down most of the neutrons from the fission energies up to 2.0 Mev to ther- mal energies of < 1 ev. The lead absorbs most of the fission gamma radiation. The resulting flux consists mainly of thermal neutrons, and has a magnitude of ~ 3 x 10 8 neutrons/ cm 2 / sec at the power level used. Thermal flux calibrations were made by irradiating copper wires at the points of exposure, both with and without cadmium shields. The thermal neutron fluxes were evaluated from the induced cop- per activity by T. Arnette, according to the usual procedure by the use of the formulas developed by H. Jonts. Fast neutron fluxes were obtained in the same tunnel by using the "fast neutron cart." This cart consists of an aluminum framework holding 42 kg of uranium slugs shielded above by four inches of lead. The thermal neutrons in the tunnel cause fission in the uranium, liberating fast nei trons of fission energies. The lead cuts down fission gamma radiation to a small value. The exposure com- partment is placed over the lead and is further shielded by sheets of boron-carbide impregnated lu- cite. The lucite contains ~ 0.324 g of boron per cm 2 , which if uniformly distributed would cut down the thermal neutron flux by a factor of ~ 2 x 10 5 , making thermal flux inside the compartment neg- ligible. Fast neutron fluxes were measured in n-units by means of a Victoreen r-meter, standardized with the r-meter at Berkeley. The n-unit is the flux which will produce the same discharge in this meter as that produced by one r of gamma radiation. The tracks were counted under a microscope, using a 97X objective and a 10X eyepiece. The diameter of a field was 0.15 mm, giving an area of 1.768 x 10~*cm 2 /field. This is equivalent to 5.66 x 10 3 fields per cm 2 . Emulsion thickness was ~ 40/i, and the depth of focus of the above-mentioned lens system was ~ 4/i, so that the objective had to be shifted vertically to count all the tracks in a field. Three or more grains in line were considered to constitute a track, although it is probable that protons of low energy produce tracks of but two grains. These latter, however, cannot be recognized with any degree of certainty and so were disregarded. There is also sometimes some question as to whether a 3-grain track is a true track or only a random configuration of fog particles. RESULTS AND DISCUSSION Qualitative Exposure of films with and without cadmium jackets to 8.66 x 10 9 thermal neutrons/ cm 2 (as de- termined with copper wires) showed that cadmium cut down track density by a factor of 4. Doubling the flux gave essentially the same ratio. Exposure of films in the "fast cart," with and without cadmium jackets, gave 5% fewer tracks in the shielded films. These tests were designed to determine whether thermal neutrons produced tracks. Since cadmium has a resonance at ~ 0.18 ev, and does not transmit neutrons appreciably below 0.4 ev, the differences in track density previously noted established the fact that nitrogen cap- ture of neutrons of energies below the cadmium cutoff was a factor in track formation. Track distribution was determined in terms of the number of silver grains per track. The re- sults are given in Table 1. The films exposed to thermal neutrons with a cadmium jacket had so much gamma fog from ra- diative capture of neutrons by the cadmium that this determination could not be made. The foregoing data show that the percentage of tracks of 5 grains or less is only slightly higher for thermal neutrons than for fast, and that there is a slightly higher percentage of tracks of six or more grains due to fast neutrons. Assuming that the number of silver grains developed is a function of the energy of the proton, one might expect that there should be a higher average energy to the re- coil protons. However, since the energy of the recoil protons is also a function of the impact angle at I MDDC - 890 [3 Table 1. Track distribution based on silver grains per track. 3-grain 4-grain 5-grain 6-grain > 6-grain (%) {%) (%) (%) (%) F 50.9 25.8 11.7 6.1 5.6 F c 48.0 27.8 12.6 6.2 5.6 Th 50.2 29.2 14.8 4.6 1.2 Where F indicates films exposed to fast neutrons without a cadmium jacket, F c indicates films exposed to fast neutrons wilh a cadmium jacket, and Th indicates films exposed to thermal neutrons without a cadmium jacket. which collision occurs, it can be seen that the proton energy may vary from near zero, for a glancing blow, to the full energy of the incident neutron, for a head-on collision. From these considerations it is obvious that track length (in terms of number of silver grains) cannot be a criterion of whether the proton-producing neutrons were thermal or fast. Exposures were also made of a nitrogen-free film, which had been made up for J. Floyd by East- man Kodak Co. This film proved to be so highly gamma-sensitive that cadmium-jacketed films were fogged too much to evaluate track densities. Investigation of this film was discontinued. Calibration for Thermal Neutrons Films were calibrated for track frequency due to thermal neutrons by making a series of ex- posures in the slow neutron tunnel, using 15-second increments. A film was run in and out of the exposure end of the tunnel to determine the number of tracks formed by the neutron flux gathered en route. A correction was made on each subsequent determination by subtracting this "background" value from the observed track density. Two films were exposed for each time interval, and 25 fields counted on each film, making a total of 50 fields counted for each exposure. Table 2 shows the results of these counts. Calculations were performed on these data to determine the number of neutrons causing a track in the film. The figure used to calculate total flux is 2076 thermal neutrons/ cm 2 / sec. This is the result of several determinations with copper wire as previously mentioned and agrees with numerous similar determinations made by the Biology Group. Since the area of each field is 1.768 x lO -4 cm 2 , the area of 50 fields is 8.84 x 10" 3 cm 2 , so that dividing a track value for 50 fields by this figure gives tracks/ cm 2 . Table 3 shows the results of these determinations. These figures give an average value of (3.01+ 0.09) x 10 s thermal neutrons/ track. Theoretical expectancy of thermal neutrons/ track was calculated, using the results of an analysis of the emulsion made by the Chemistry Division for L. B. Borst some time ago. 1 This analysis showed the N:C:0:H ratio to be 1:2:2:5, or the same as that for glycine. Actually, gelatin has a more com- plex structure. However, using the values for glycine: sp gr = 1.6 mol wt = 25.07 ratio of nitrogen atoms to total number of atoms = 0.1 thickness of emulsion = •» 40/j. 12-227-p3-bi 4] MDDC - 890 Table 2. Tracks due to thermal neutrons. Exposure time (sec) Tracks/50 fields Tracks/field Tracks/field less background 1.04 ' — 2.46 1.42 4.86 3.82 6.40 5.36 9.06 8.02 10.78 9.74 12.10 11.06 13.70 12.66 14.80 13.76 0.0 (background) 52 15 123 30 243 45 320 60 453 75 539 90 605 105 685 120 740 Table 3. Determination of number of thermal neutrons per track. Exposure time Tracks/50 fields Tracks/cm 2 Thermal Thermal neu- (sec) (backgrd subtracted) neutrons/cm 2 trons/track 15 71 8.02 x 10 s 3.11 x 10 9 3.88 x 10 5 30 191 2.16 x 10* 6.22 x 10 8 2.88 x 10 s 45 268 3.03 x 10* 9.32 x 10" 3.08 x 10" 60 401 4.54 x 10* 1.25 x 10 10 2.20 x10 s 75 487 5.51 x 10* 1.56 x 10 10 2.83 x 10 5 90 553 6.26 x 10* 1.87 x 10 10 2.94 x 10 5 105 633 7.05 x 10* 2.18 x 10 10 3.09 x 10 5 120 688 7.78 x 10* 2.49 x 10 10 3.20 x 10" Then Dm =-!£-£ — x rtj x Avogadro's number x thickness mol wt where Dm = density of nitrogen atoms and r N = ratio of nitrogen atoms to total. Substituting numerical values: D N = 1 - 6 x 0.1 x 6.02 x 0.10 23 x 4 x 10" 3 75.07 5.12 x 10 18 nitrogen atoms/cm 2 . la-317-p4-ti> MDDC - 890 [5 Proton flux due to N"(n,p)C" is P N =aa N xD N xn, where Pjj is the proton flux produced in the above reaction, a a N is the thermal capture cross -section of nitrogen, and n is the thermal-neutron flux. Substituting numerical values: P N = 1.7 x 10 _M x 5.12 x 10 18 x 9.32 x 10 9 * = 8.11 x 10 4 protons/cm 2 in 45 sec. A correction must be applied here. The chemical analysis 1 previously mentioned showed the nitrogen content of the emulsion to be 14 per cent, instead of the 18.5 per cent which it would have been if the emulsion had been pure gelatine. A part of the emulsion comprises silver halides, and It has been determined 2 that at a relative humidity of 50 per cent (which prevails in this area) gelatin absorbs 20 per cent of its weight of water. Judging by the nitrogen percentages, it appears that 75 per cent of the emulsion is gelatin. This correction brings proton expectancy down to 6.10 x 10* pro- tons/cm 2 /^ sec. Calculation of the corresponding values for each exposure, checked against the figures in column three of Table 3, shows that 53 per cent of the protons formed recognizable tracks of 3 grains or more. The rest probably produced 2-grain tracks, but, since these are difficult to identify, it was considered advantageous to disregard them for purposes of personnel monitoring. Calibration for Fast Neutrons Calibrations for fast neutrons were made in the "fast cart." Increments of 0.2 n-units were used. Background correction was determined in like manner as for thermals. The results of the exposures appear in Table 4. Table 4. Tracks due to fast neutrons. Exposure — n (background) Tracks/50 fields Tracks / field Tracks/field (Background subtracted) 1.36 — 2.78 1.42 4.06 2.70 5.86 4.50 6.94 5.58 7.80 6.44 10.40 9.04 12.20 10.84 13.50 12.14 14.12 13.76 16.84 15.48 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 68 139 203 293 387 390 520 610 675 756 842 ♦Value for 45-sec exposure. 13-337-pb-bu g i MDDC - 890 Using the foregoing data, calculations were made to determine the number of fast neutrons indi- cated by each track. Neutron exposures were calculated from n-units. As previously mentioned, an n-unit is the fast neutron exposure which causes a discharge in the Victoreen r-meter equivalent to that caused by one r of gamma radiation. P. C. Aebersold has estimated that tissue absorbs about 205 ergs/g when exposed to one n-unit of fast neutrons. Since one rep of radiation is, by definition, that amount which loses 83 ergs/g of tissue, one n-unit is approximately 2.5 rep. C. C. Gamertsf elder has calculated 3 that 2.84 x 10 8 neutrons of 2 Mev will lose 83 ergs/g of tis- sue, or 3.83 x 10 8 neutrons of 1 Mev will lose the same amount. Consequently, to produce one n-unit requires 7.10 x 10 8 2-Mev neutrons/cm 2 , or 9.58 x 10 8 1-Mev neutrons/cm 2 . Since the mean value of the fission energy spectrum is nearer 1 Mev than 2, the latter value was used in the calculations. A correction is required for the fission gamma radiation which penetrates the four-inch lead shield. This has been estimated to account for 12 per cent of the r-meter readings. With this correc- tion applied, an n-unit, as measured on the r-meter, indicates 8.42 x 10 s neutrons/cm 2 . The results of the determinations follow. Table 5. Determination of number of fast neutrons per track. Tracks/50 fields Exposure (n) (Background subtracted) Tracks/cm 2 Neutrons/cm 2 Neutrons/track 0.2 71 8.02 xlO 3 1.69 x 10 8 2.11 x 10 4 0.4 135 1.53 x 10 4 3.37 x 10 8 2.20x10* 0.6 • 225 2.54 x 10 4 5.06 x10 s 1.99 x 10 4 0.8 239 3.15 x 10 4 6.75 x10 s 2.14x10* 1.0 321 3.63 x 10 4 8.42 x 10 8 2.32 x 10 4 1.2 452 5.11 x 10 4 1.01 x 10 9 1.97 x 10 4 1.4 542 6.13 x 10 4 1.18 x 10 9 1.93x10* 1.6 607 6.87x10* 1.35 x10 s 1.96 x 10 4 1.8 688 7.78 x 10 4 1.52 x10 s 1.97 x 10 4 2.0 774 8.76 x 10 4 1.69 x10 s 1.93 x 10 4 These figures give an average value of (2.05 ± .03) x 10 4 fast neutrons/track. A theoretical estimate was then made of the expectancy of tracks due to fast neutrons in the film. The chemical analysis previously mentioned 1 was used as a basis for the estimate of probable com- position. Using glycine as the representative compound, as above, sp gr = 1.6 mol wt = 75.07 ratio of H atoms to total number = 0.5 thickness is ~ 40ji Then: Djj = sp gr/mol wt x rjj x Avogadro's number x thickness where D H = density of hydrogen atoms and r H = ratio of H atoms to total. 13-237-p6-bu MDDC - 890 [ 7 Substituting numerical values: \ D H = 1.6/75.07 x 0.5 x 6.02 x 10 23 x 4 x 10" 3 = 2.56 x 10 19 H atoms/cm 2 . Since only 75 per cent of the emulsion is gelatin, the corrected value of H atom density is 1.92 x 10 19 . There is an additional hydrogen content of the emulsion due to the approximately 20 per cent water content of the emulsion. 2 The number of water molecules per unit weight of the emulsion is in inverse ratio to glycine molecules as their molecular weights, and their relative weights are as their percentages. Then, since there are 3.84 x 10 18 glycine molecules/cm 2 , the number of water mole- cules/cm 2 is 3 - 84xl ° J8x f§ X Tff = 4-27x10" and there are twice as many H atoms, or 8.54 x 10 18 H atoms/cm 2 , due to water content. Adding this to the H atom density due to glycine, the result is D H = 2.77 x 10 19 H atoms/cm 2 of emulsion. Proton flux, P = cK x Djj x n where