Journal of Research of the National Burea u of Sta nda rds Vol. 62, No.1, January 1959 Resea rch Paper 2924 Preparation of New Solution Standards of W. B. Mann, L. L. Stockmann, W . J. Youden, A. Schwebel, P. A. Mullen, and S. B. Garfinkel

New radium-solut standards have been prepa red in t he r a nges of 10 micrograms a nd also 10- 9 and 10- 11 gram of r adium element. These h a ve been compa red with thc Nationa l Bureau of Standard's 1940 and 1947 series of radium-solut ion standards and, as a result of t hese compar isons, it has been found t hat t he 1940 10- 9 a nd 10- 11 -gram solut ion standa rds contained som e 2 to 3 percent more r adium element t han cer tified. It has been sho wn t hat t his difference pro ba bly arose in t he dilutio n of t he 1940 standa rds. 1. Introduction out using the NBS radiation balance [4 , 5) . These measurements consisted of three m easurem ents of R adium-solution standards have previously been the rate of energy emission from the new radium prepared at. the National Bureau of Standards in 1940 som'ce alone and also one triad of m easurem ents and 1947. The 1940 series consisted of standards in [1 , 2) of the new source relative to both primary the microgram range, ranging from 0.1 to 100.0 J.1.g of standards (No. XIV and XV). The r esults of these Tadium element in 5 ml of solution, and standards for measurem ents are shown in table 1. calibration consisting of 10- 9 and 10- 11 g of radium elemE'nt in 100 ml of the radium-salt and car­ T ABLE l. Radiation-balance results Jor comparison oj new m dium source with the United States primary radium standards I rier solution. The 1947 series consisted only of micro­ 'gram standards ranging from 0.1 to 100.0 J.1. g of ra­ R ate of energy emission in micro· Milligrams I< :l.ium element in 5 ml of solution of the radium watts from- of and nitric acid acting as carrier. D ate of measurement element in new source R ecently the stock of 1O- 9_g r adium-solution stand­ No. XIV No. XV New source I ards became so depleted that it was necessary to prepare a new set of standards which has been des­ October 23, 1956. _.. _.. 914.8 6.108 November 1, 1956 ____ . 914.8 6.108 ignated as the 1957 series of tandards and which November 8,1956 ______914.6 6.107 consists chiefly of 10- 9 and lO- 11 _g standards with a November 28, 1956 .. __ . 5728.0 3000.3 914.1 6. J04 few microgram standards which were prepared for comparison purposes. A new set of "blank solutions" In calcula ting the values shown in table 1 a cor­ was also prepared consisting of 100-ml samplcs con­ rection was m ade for tbe growth of polonium-210 I taining 0.2 percent by weight of B aCb ·2H 20 . in the national standards sin ce June 1934. The mean value of the rate of . energy emission from the 2. Radium Calibration new source is 914.6 J.1.W which corresponds to 6.107 mg of radium elemen t as of N ovember 1956. The A sample of radium containing approxi­ gamma-ray comparison, carried out with the gold­ m ately 10 mg of radium elem ent was retUTned to leaf electroscope, gave an average value from twelve the Radium Chemical Company for a r eseparation measurem en ts equal to 6.08 mg of radium element. from radium D and Its products and for recrystalli- zation. It was requested that the radium salt should Ibe crystallized in such a manner that the grain size 3. Preparation of the New Radium-Solution ,would be of the same approximate dimensions as Standards t hose in the Honigschmid radium standards (which wer e also radium chloride) and that the radium salt The 6.10 7-mg radium source was now completely Is hould be enclosed in a glass tube of about the same shattered at the bottom of a 5-liter thick-bot tomed dimensions (length, diameter, and wall thickness) as glass bottle under 3.052 liters of carrier solution, t he tubes used by H oni gschmid. It would then be determined by weighing and consisting of 0.2-per­ possible to compare this radium SOUTce wi th the two cent BaCl· 2H20 plus 5-percen t Hel, b y imparting a sharp blow to the glass tube by means of a specially United States primary radium standards [1 ,2], 1 using t he BS gold-leaf electroscope [3], without making constructed glass rod with a thickened and elongated any absorption corrections. In such a comparison end which was struck at its other end with a hammer. the sources are supported horizontally and then By this procedure the master solution of radium and carrier , with a concentration of radium elem ent of gen tly tapped so as to spread the grains of salt uni­ 6 formly along the glass t ubes. 2.001 X 10- glml, was prepared. Wbile the radium SOUTce was compared in this The dilutions that were made from this master manner with the two primary standards, this com­ solution are shown diagrammatically in fi gUTe 1. parison was only treated as confirmation for a series These dilutions were carried by two independent of microcalorimetric comparisons which were carried routes, designated as A and B , in order to check the aCCUTacy of dilution. The master solution as well

J F igures in b rackets in dicate the literature references at the end of t his paper. as all subsequent dilutions thereof were thoroughly 21 mixed by agitation before aliquots were removed. parison with the 1940 series of 1O-9_g and 10- 11 -g All glassware used was carefully calibrated. radium-solution standards. The nominal values of First of all two lO-ml aliquots were each diluted the dilutions shown by each rou te were: 2 X 10- 7g lml, to 100 ml in a 100-ml volumetric flask using carrier 2 X 10- 9g/ml, 1 X lO- llg/ml, and 1X 10- 13g/ml. solution. Following this, eight 5-ml aliquots were Four of the eight 1957 10-/-Lg radium standards pipetted into glass ampoules and flame-sealed . were now compared in the NBS 41TI'-ionization cham · These eight ampoules each containing 10.163 /-L g of bel' [6 , 7] with four 10-/-Lg radium-solution standards radium-226 per 5.079 ml of solution were set aside of the 1947 series and were found to agree with the for comparison with the microgram range of both 1947 values to within the ± 1 perce nt es timated ac­ the 1940 and 1947 standards by means of the NBS curacy of the 1947 standards. Subsequently three 41T1'- ioniza tion chamber. of the 1940 series of 10-/-Lg radium-solution standards, At this point the r emainder of the master solution the stock of which had been believed to be exhausted, was siphoned off into two 2,500-ml volumetric were found ancl compared with three of the 10-/-Lg flasks and flam e sealed for future possible use. The standards of t be 1947 series and three of the 1957 remaining small volume containing the fragments of series . the glass tube was check ed and found to contain no Due to the quite large calibration correction of more radium per milliliter than one of the 10-fLg the 5-ml pipet (the volumc was equal to 5.079 ml) samples. tbe nominal 1957 10-/-Lg radium-solution standards The further dilutions along routes A and B were have an actual radium content of 10.163 /-Lg. The carried out as shown ill fi gure 1 and gave, by each comparisons of the 1957 standards with the 1947 rou te, fif ty 1O-9_g and fif ty lO-ll_g radium 100-ml standards were carried ou t in April and August solution standards. Of these the first, twenty-fifth, 1957, wbile that of the 1957 with the 1940 10-/-Lg and fiftieth 1O-9_g and 10- 1l-g ampoules in both standards was carried out in August 1957. route A and route B were reserved for later com- The certified values of both the 1940 and 1947

MASTER SOLU TIO N: 6.107 mg Ro - 226 IN 30S2 ml OF CARRIER SOLUTION 10.2% BoCI 2H2 0 + S% HCI ) '2.001 X 10"6 GRAM Ro per ml

MASTER SOLUTION 30S2ml

10 mi 10mi

~7~:~,~~.~~,~"~'~10mi STAN DARD AMPOULES FOR 10 ml 4,,1' COMPARISONS ·10.16i X 1

FIFTY IDO - ml SAMPLE S FIFTY IDO-ml SAMPLES '10" GRAM Ro EACH . NOS. ' IO"GRAM Ro EACH . NOS. IA.2SA AND SOA RESERVED IB.2SB AND SOB RE SER V[O FOR COMPARISON WIT H FOR COMPARISON WITH 1940 STANDARDS 1940 STANDARDS '"o

~0- W 0:

2000 m l FIFTY 100-m! SAMPLES FIFTY 100-ml SAMPLE S '" la-II GRAM Ro EACH ' 10" GR AM Ro EACH

FIGU RE 1. Dilution scheme for the prepnration oj 1957 10- 9-g and 10- 1.-(/ radium standards. 22 L "10-Mg" radium-solution standards ~re ] 0 Mg as of four microcalometric comparisons of the 6-mg ra­ August 1940 and lV' ay 1947, respectively. dium preparation with the national radium standards In terms of the 1940 10-Mo· radium standards the was such as to give a standard deviation of the 1957 "10-Mg" standards were found to co ntai~ 10.1 50 average of 0.02 percent. ± 0.0014 Mg of radium in August 1957 . This value is in terms of different 1940 ampoules and the agree­ 4. Comparison of the 1940 and 1957 10- 9_g ment is well within the almost ± 1 percent uncer­ Radium-Solution Standards by the Method taintyof the values of the 1940 standards. of Radon Analysis In terms of the 1947 10-Mg radium standards the 1957 "10-jlg" standards were found, in the August The method of radon analysis in use at the N a­ 1957 measurements in the NBS 47r'Y-ionization cham­ tiona,} Bureau of Standards is essentially the method ber to have a radium content of ]0.230 Mg in con­ described by Curtiss and Davis (8, 9], but wiLh the tra~t to the calibrated value oJ 10.163 Mg. This reflux condensers for de-emanation of the radium value is again within the almost ± 1 percent un­ solution standards replaced by wash bottles with certainty of the 1947 standards. sintered-glass filters as described by Harding, Finally a number of the 20-Mg and 50-Mg standards Schwebel, and Stockmann (10J. The radon is re­ of the 1947 series were compared with those of the moved from these solutions by m eans of a fine stream 1940 series. TIle complete series of 47r'Y-ionization­ of nitrogen bubbles passing through the solutions chamber measurements is summarized in figure 2 from the sintered-glass filters. In order to confi..rm where, on the left-hand side the results of the com­ the results obtained with this method of de-emanatIOn parisons of the 1940 to 1947 standards are shown. the reflux condensers were, however, reinstalled for a In this case bo th the 1940 and ]947 standards are final set of comparisons. certified as 10, 20, and 50 Mg. The small errors in­ Because of the somewhat lower precision of the clicated are those to be associated with the 47r'Y­ radon method of analysis, as compared with the ionization-chamber measurements while the larger gamma-ray measurements already described, a most errors are those inherent in both the 1940 and 1947 exhaustive series of intercomparisons between the standards. Within these latter limits there is no 1940 and 1957 series of standards was carried out by deviation from unity. the radon method. On the right-hand side the ratio of (,he four 1957 As indicated in figure 1 ampoules lA, 25A, 50A, 1B, to seven 1947 10-Mg standards is shown, afler cor·· 25B, and 50B were selected hom the 1957 dilutions reeling for the volume of the 5.079-mlpipet ( us~ d in for comparative measurements. the 1957 series) to the equivalent of the 5 ml (1. c., Four 100-ml ampoules, designated as A, B, C, and correcting to 10.000 Mg instead of 10.163 Mg for the D were available hom the 1940 1O -9_g radium­ 1957 series). After this correc tion has been made soiution standards and were used to compare with the ratio of the 1947 series (certified simply as 10 Mg) the six 1957 ampoules. to the 1957 series should be unity. Once again the The 1940 10- 9-g standards were certified just as deviation from unity (1.0054 ± 0.0004 , the error of 10- 9g of radium element. The certified value for the the 47r'Y-ionization-chamber measurement) is well 1957 standards is O.999 X 10- 9gofradium . The rat,io of within the 0.8 percent "uncertainty" certified in the activities of the 1940 and 1957 standards was now de­ case of the 1947 series of radium-solution standards termined by using each in tum to calibrate eight NBS above, without even considering the errors inherent. alpha-particlc-pulse-ionization chambers [8, 9J that in the 1957 series which are estimated to total about are routinely used for radon assays. The results ± 0.1 percent or 0.2 percent. The actual precision in Were expressed in terms of the ratio of the activity of the 1957 radium-solution standards, as determined by the calibration (in terms of counts per unit of RATIO OF 1940 TO 1947 STANDARDS RATIO OF 195710 1947 STANDARDS radon) of any given chamber to that of tl?-c 1940 IN TERMS OF ACTIVITY IN TER MS OF ACTIV ITY 1.020 radium-solution standards corrected for radmm de­ cay, using the same chamber, and are shown in table 2 (a). 1.010 r-... TABLE 2 (a) . Comparison of 1957 and 1940 10- 9-g standm·ds (T he entries are the resul t of dividing the activities of the J95i standard by th os() of the 1940 standards) 1.000

195i Standards 1940 Standards .990 IA 25.'\. 50A IB 25 B 50B ------A ____ . __ . _ 0.9iO 0. 954 0.9il O.9iO 0. 9i5 0.9i4 B ___ .. __ . _ .08i . 958 . 980 . 9i4 . 98i .964 .980 C __ . ____ ._ .982 .969 .983 . 980 . 962 . 964 D ______" 101<9 . 983 . 97i . 965 .980 .9i6 . 980 3 b~I

23 MASTER SOLUTION CONSISTING The average ratio of 0.9740 derived from the results OF SIX 10-MICROGRAM RADIUM of table 2 (a) for the 1957 and 1940 standards indi­ GAMMA-RAY STAN DARDS EACH cates that the certified radium content of the 1940 CONTAIN ING 5 ml OF SOLUTION standards was low by 2.6 percent. Subsequently eight reflux condensers were re­ installed for de-emanation of the standard solutions by boiling. The radon was fed from two of these reflux condensers into two of the eight alpha-particle­ ionization chambers that had been used to obtain l oml~ the results in table 2 (a). The values for the ratio 100 oml 100 obtained for the 1957/1940 activities (again obtained m l ml from the chamber calibrations) using the reflux con­ densers, and again corrected for decay, are shown in cVIOml lomi table 2 (b). These last results were not, however, used in calculating the average ratio of 0.9740 as they were not systematically determined for every combination of the 1957 standards (lA, 25A, 50A, 1B, 25B, and 50B) on the one hand and the 1940 stand­ ards (A, B , C, and D) on the other. The refiux­ condenser measurements were merely confirmatory. lomi

TABLE 2 (b). Comparison of 1957 and 1940 10-9-g standards using reflux condensers for de-emanation

Ratio of activities of 1957 to 1940 standards

1957 Standards 1940 Standards FIVE IOO-ml SAMPLES OF FIVE 100-ml SAMPLES OF 50A 25B 50B 109 GRAM OF RADIUM, 109 GRAM OF RADIUM, NUMBERED l,lI,m,Ill NUMBERED"ll, Jll,W,'lZlIr AND IX RESPECTIVELY AND:X: RESPECTIVELY L ______{ 0. 978 0. 977 --- .. . 978 . 975 0.984 B ______9 . 987 . 986 ~ ...... FIGURE 3. Dilution scheme fo r the pnparation of 10- - (1 D ______{ . 986 . 985 . 989 samples from 1947 radium gamma-ray standards . . 9&1 . 980 -- .. -

Average ratio 1957/ 1940 =0.9823 and for the 1947 and 1957 1O- 9_g samples and stand­ ards with the results shown in tables 3 (a) and 4 (a), the appropriate corrections again being made for This discrepancy of 2.6 percent was so large that radium decay. The sintered-glass-filter method it was considered desirable to check the 1940 and 1957 of de-emanation was used in these measurements standards against the 1947 standards to try to de­ and also the same eight alpha-particle-ionization termine which was the more likely to be in error. chambers as were used to obtain the 1957/1940 Unfortunately, however, the 1947 series consists ratios shown in table 2 (a). In view, however, of only of standards in the range of microgram quanti­ the greater numbers of 1940 al?-d 1957 .10- 9-g ties of radium element. It was, therefore, necessary standards and 1947 10- 9-g samples mvolved It was to carry out a dilution of 1947-microgram standards not possible to compare every standard solution vyith to the 1O- 9_g level. This was, however, accepted as every 1947 1O- 9_g sample. A pattern of companson an additional check on the accuracy of our dilution. was devised as indicated in tables 3 (a) and 4 (a) The dilution scheme is shown in figure 3, the initial to give a m~ximum number of interchecks wit~out master solution consisting of six 10-,ug radium­ taking every possible c?mbination.

TABLE 4 (b ).-Comparison of 1947 10- g-g samples and 1957 As a further check between the 1940 and 1957 g 10- -g standards using reflux condensers for de-emanation series of standa,rds three 1940 10- 11-g radium-solution standards of 100 ml, certified as containing l.025 X Ratio of activities of 1947 to 1957 standards 10- 11g and designated as 11, 12, and 13, were com­ pared with three 1957 lO- ll_g radium-solution lO- '-g solutions made from six 1947 gamma-ray standards standards of 100 ml, designated as 21, 22, and 23 1957 Sta ndards and found to contain l.001 X 10- 11g, by the method I II V VIII of radon analysis using alpha-particle-ionization chambers 7 and 12 and de-emanating by boiling in 5OA. ______- ______{ 0. 988 0.990 0. 984 0. 986 the reflux condensers. . 993 ------. -- - -- 25B.. ______{ 0. 996 } 0. 983 0. 988 0.989 In view of the longer collection times that were 5OB ______- ______. 989 . 993 . 985 ----- involved and the fact that the readings were only 0. 987 0. 984 - --- - 0. 980 some ten times background the results took much

Average ratio 1947/1957 = 0.9885 2 The reciprocals of the ratios as determined from the original values are shown - for convenience in computation, should they be desired. 25 longer to obtain and it was not, therefore, possible radon mothod for measurements which are, in this to carry out as exhaustive comparisons as with the case, of tbe order of one-tenth background. The 1O-9_g standards. The results, after correction for average value obtained, however, for the 1940 blank radium declty, are shown, however, in table 5. The solutions wa s 0.18 X 10- 1Zg of radium per 100 ml as average ratio of 0.986 (1957/1940), with a standard compared with the certified value of 0.25 X 1O-12g of deviation of the average of 0.020, is in fairly close radium per 100 ml. agreement with the value of 0.9774 (1957 /1940) for It appears that the 1940 lO- lI_g radium-solu­ the adjusted ratio for the 10- 9-g radium-solution tion standards which are certified as containing standards. It must, however , be borne in mind that 1.025 X 10- 11 g of radium were derived from the dilu­ additional errors are introduced at such low concen­ tion of the 10- 9-g radium solution, certified as con­ trations as lO- ll _g radium per 100 ml by uncertain­ taining 10- 9g of radium per 100 ml, with the blank ties in the radium content of the diluting carrier solution which was said to contain 0.025 X lO - llg of solu tion, as will be appaTent from measurements radium per 100 ml. If this last figure were obtained, made on such solutions which will now b e described. however, by the radon method it would appear to be no more reliable than that of 0.18 X 10- 12g of T ABLE 5. Comparison of 1959 and 1940 10- Jl -g standards 1lsing reflux condensers for de-emanati on radium per 100 ml recently determined. Under these circumstances the value of 0.986 11 R ad ium content in u nits of 10-l1g obtained for the ratio of t he 1957/1940 10- _g Standard Ko." radium solution standards cannot be said to diffor Chamher No.7 Chamber No. 12 significantl:v from that of 0.977 obtained for the ratio of the 1957/1940 1O-9_g radium-solution 11 ____ . ______1. 040 standards. 21 ______. ____ . _____ . ______l. 032 12 ______. ____ _ 1.001 7. Summary 22 ______l. 016 11 . ______1. 047 From t he measurements on the 1940 and 1957 21. ______. __ 1. 022 1O -9_g radium-solution standards, which are co n­ 12 ______1.007 13 ______. __ 1. 032 fi['m od by those of the lO - ll_g series, it would appear 23 ______. __ 0.990 13 ______. ______that there is a n error of abou t 2.6 perce nt in tho 1. 018 1940 series of 1O -9_g and lO - ll _g radium-solution 23 ______1. 012 22______0.987 standards. As no corresponding di scr epanc~~ has bee n observed in the miorogram serif'S of standards Average ratio 19 57 / 1 940 ~ 0.986 it is assumed that the error is one which occulTod Standard deviation ~ 0.020 in t he dilution down Lo 1O -9g a nd lO- llg per 100 ml ill 1940. a 1940 Stand ards deSignated as 11, 12, and 13 ; 1957 Standards deSignated as 21, 22, a nd 23. 8. References [1] T. I . Davenport, C. C. McCraven, W . B. Mann, and 6. Determination of the Radium Content of C. C. Smith, Comp 9rison of four nationitl radium the Carrier Solution standards, Part 1. Experimental procedures and re­ sults; W. S. Connor and W. J. Y ouden, Part 2. The carrier solution used in the dilutions, shown Stat i st i c.~ l procedures and survey, J. R esearch NBS 53, 267 (1954) RP2544. schematically in figures 1 and 3, consisted of 0.2 [2] T . P. Loftus, W. B. Mann, L. Paolella, L. L . Stock­ percent by weight of BaC12·2H20 . In order to mann, and 'vV. J. Youden, Comp.