Radium and Thorium X in Some Potable Waters (Acta Radiologica, Vol 43, January 1955) (Partial Document)

Page 1

CONDITA AGO ST A FOR S SELL EDIT A IUS SU SOCIET ATUM RADIOLOGIC ARUM DANI A E FEN NIA E,NOR VE GIA EET SUE CIA E RADIUM AND THORIUM X I N SOME VOL. 4 3 JANUARY 195 5 FASC. 1

Page 2

A number of radium analyses have be-3 made in order to obtain data on natural levels of radioactivity and -0 help locate potable water supplies having an unusually high radiuzzl content as part of a current research program on the toxicity of radi a i. Several findings of general interest are described in this paper. These niny be summarized as follows: 1. The concentration of radium in surface water or shallow wells in the Chicago area is low - less than 0.1 I I ~ Cper liter. 2 . Water from wells which penetrate into the Galesville formation at a depth of about 1,500 feet or more near Chicago is found to contain from 3 to 8 ppc of Razz6 per liter and about the same amount of ThX (3.64-d Razz'). This is believed to be the highest authenticated concentration of radium found to date for an\- drinking water in the United 3. Practically none of the immediate parent of ThX, 1.90-y RdTh, is present in the deep well waters, so the ThX decays after sampling. 4. Drinking water from Frankfurt am Main, Germany, was found to contain less than 0 . 3 ,upc Ra per liter as compared t o more than 300 ppc per liter previously reported (8). Nuch higher values for the radium content of some spring and mine waters have been reported (10, 7). However, i t is recognized that some of the earlier values reported as "radium" were in reality for radon, the activity of which is often several orders of magnitude greater than that of radium in the same water eample. An extreme case of the disparity of radium and radon concentrations in a single water source was reported by KL-RODAand YOICOYA;\IAwho found a Masutomi spring (designated as A49) in Japan to contain 1 . 3 3 x los p p c Rn per liter Submitted for 1 ublication, 28 April 1954.

Page 3

and only 8 / C U C Ra per liter (9). The highest radium content reported by KURODAand YOKOYAMAwas 76 y,uc per liter for Masutomi Spring As. The highest concentration HURSH( 3 ) found recently in a large number of samples of drinking water from the United States was less than 2 ,u,uc Ra per liter. Values of 190 to 480 < 10-12 grams Wa per liter for the drinking water of Barcelona, Spain, \vert. given by LOVEwho did not distinguish beh-een radon and radium in his review paper (IO). However, CANALSCRAMP(3) actually had reported his findings on the Barcelona water as radon, and stated that the radium content was too low t o be measured with his apparatus. The necessity of distinguishing between ThX and long lived Ra”a in analyses for Ra has received little attention, because most analyses for radium are done by collecting radon, and the presence of small amounts of ThX xvould have a negligible effect on the results. If the ratio of T h S t o Razz‘were high, or if radium analyses were made by isolation and nieasurenient of radium then an appreciable error could be made in reporting results as The following values (in p p c per liter) reported (9) for Masutomi Spring B7 illustrate a situation in which analytical techniques must clearly distinguish between the radioactive nuclides associated with .‘radium”: The radium content of public water supplies is usually niuch lower. i Analytical Procedures. The chemical procedure used to isolate radium was essentially that developed by AVES and co-workers (1) and RCSSELL,LESKOand SCHUBERT(12). The sample (usually 1 liter) of water was made 0.1 31 in HNO,. evaporated t o dryness. and baked t o eliminate excess acid. The residue was taken up with 0.1 31 HNO, and transferred to a 50 ml centrifuge tube. If any insoluble matter remained it was centrifuged off and the clear supernatant retained. -4 volume reduction step was made b y precipitating PbSO, which carried radium from the solution. After centrifuging, the precipitate nas dissolved in a mixture containing six parts concentrated HC1 and one part ethyl ether. Then 4.0 mg B a + + carrier was added t o precipitate BaC1, which carried down the radium. After centrifuging, the precipitate was dissolved with a few drops of 0.1 81 HNO, and the solution transferred t o a platinum plate and evaporated t o dryness after adding 4 drops 0.5 M H,SO, t o form a n adherent precipitate of BaSO,. After flaming. the plate was ready for counting. The activity n-as measured by counting alpha particles in a windowless proportional counter which had a background of about 0.05 counts per minute (cpm). The tlisinteprutions per minute (dpm) were calculated from the formula: the counting rate of standmd radium anniplea with R:i cirricr .l[111t31! . I I ~ I ~\ v i t h nu carriel-

Page 4

3 too IOW to G : percent W ~ R found by analysis of aliquots of a solution of radium standardized by ;I> they grow in. Aiccortiingly,the counting r;ltes obsjervetl had t o b e corrected t o the true wtlium act,ivity b y iisc of tht: growth f w t o r for Ra"'6. R, c:1lcnlate(l as follows: where the tlecay constant. for Rrl""? is O.llt)ii > hr-1antl t is the riunibcr of hours after isolation of ratliuni. Fnrthrrniore. a s will be shown later. T h S was also present in marly of t.he samples. I n this c;tse 54.5-s Tri arid O.i(;-s Th-1 grew i n so rapitlly that three alpha tlisintegrations were obtained f o r every tlisintegration of T h S when counting was started. Another alpha-emitting tiexendant. ThC. grows in more slowly so that the overall growth When both Ra't6 m icl Th S were prtwnt the aniounts of each were calculated b y counting the sample twice: once immediately after preparation of the sample and again after a period of R week or niore. The equations used are as follows: where Colll and Cox are the true counting rates for RaZz6arid ThX, respectively, a t the time radium was isolated. C, and C, are the first and second counting rates obtained, and R and X are factors taken from equations 1 and 2 for the times after isolation of radium at which C, and C, were obtrtinetl. I n practice, the samples were counted a t least three tiniea in order t o he sure t h a t the observed growth or decay corresponded to that of Radium Content of Water i n the C'hicugo Area. Samples of drinking water were obtained from communities in the vicinity of the Du Page County site of the drgonne h'ational Laboratory, twenty miles southwest of Chicago. During preliminary studies on some of these samples, the ThX isotope of radium was detected by a study of the growth curve of the alpha activity of the isolated radium. Figure 1 illustrates the growth curves obtained when separate portions of a sample of untreated water from Joliet, Illinois were analyzed 1, 7 , 9, 20, and 26 days after the sample v a s collected. The first count of each succeeding portion was lower than that of the previous portion and the growth for the later portions corresponded more closely to the theoretical for RaZz8.These data were interpreted as a result of a mixture of Ra226and ThX in which T h S was decaying because it was not being supported in the sample by

Page 5

Fig. 1. Growth curves of Ra226and ThX separated from portions of a sample of Joliet well water a t different times after sampling. -- Counting rates for a given portion. - - - First counting rate for a given portion. -'-- Decay of ThX obtained by subtraction of counting rate of Razz@ from first counting rate for a given portion. - Calculated growth for Ra226. (calculated using Equations 3 and 4) are given in Table I for the Joliet sample and for a deep well on the ANL site. The calculated values for Razz6 were constant, and the concentrations of ThX a t the collection time (Table I, column 6) were constant within experimental error when the values a t the time of analysis were corrected for decay with a 3.64day half-life. Long counting times were used in the analysis of the ANL sample, in order t o obtain better precision than that obtained for the some thought was given t o the possibility that the decay ascribed t o the former might have been due, instead, t o contamination of the BaC1, precipitate by radioactive decay products of radon present in excess of radium. This possibility was rejected for several reasons: 1 ) The concentration of radon in the Joliet sample was found t o be 2.5 x 102 p p c per liter, and, since radon is quickly boiled off, the

Page 6

RADIUY AND THORIUM S IS SOME POT.1BLE WATERS Times After CollectiotL or Siltnples 1 Elapsed time from collection of sample to isolation of radium. 2 The errors indicated are the statistical 0 . 9 error in counting or the average error 3 Concentration of ThXat time of isolation for a given portion. of duplicate determinations, whichever was the larger. activity of the daughters having an average half-life of about 30 minutes would have been reduced to less than 1 p p c by decay alone in 4 hours, the minimum time elapsed for evaporation and mounting 2) similarly, the short-lived descendants of radon would have been reduced t o negligible levels for the A S L sample because the third and fifth day analyses were made on water which had been evaporated from 5 liters to 500 ml on the day of sampling; 3) furthermore, the chemical operations used for the analysis are reportsolving of PbSO, in concentrated HC1 should have been an effective 4) no combination of RaZz2'and descendants of Rn?z? seemed likely to result in the growth curves given in Fig. 1. Direct analysis for thorium in the water from the ANL well showed that less than 0.02 p p c Th per liter was present. This confirmed the lack of RdTh indicated by the decay of ThX. Since the presence of ThX was established, calculations for Rages and ThX were made for all samples. As seen in Table 11, most of the results on water obtained near the ASL site fall into two groups: high

Page 7

1. Deep Well #LI at A\XL, 1,600 Fcet ........ 5/3/,52 15. Elgin - State Hospital, Tap .............. 20. St.Lteville - State Prizon, Center Well, 1,500' . 6 4/31 2 1.8 0.1 9.1 + 0.2 21. Stciteville- State Prison, Outside Well, 1,560'. 6/1/51 2 4.2 & 0.5 7.2 & 1.2 22. Stateville - State Prison, Ontsiile Well, 1,550'. 10/9/52 2 4.0 & 0.2 5 . 4 5 0.4 The error indicated is the statistical 0 . 9 error in counting or the average error in Pumping started for sample. replicate analyses, whichever was the larger for a given sample. values for deep wells, which penetrate into the Galesville formation a t 1,500 to 2,000 ft below ground level, and much lower values for surface water or shallow wells. It should be noted that all of the water obtained from a deep w*ell may not come from the greatest depth. because well casings are usually set only part way down the well. and water can enter from higher levels. This may explain the low values obtained for the deep well a t the Elgin State Hospital from which samples were taken soon after pumping was started. Csing a radon emanation method on samplw collected a t the same time as our samples, J. B. HURSH(a) a t the Iiriiversity of Rochester, found 6 . 5 4 ;C, 0 . 3 6 ,upc Ra per liter for the untmLtec1 Joliet sample collected on 22,4#'52and 5.79 f 0.13 p p c Ra per litfli. for the Joliet tap water

Page 8

collected on 26,5 52. These values are somewhat higher than our values of 5.4 and 4.7, respectively. The combined population of Aurora, Elnihurst, Joliet, Lernont, and Lockport is over 100,000, so these coniniunities contain a large number of people who are drinking water of a radium content appreciably higher than that conimonly encountered. It would be interesting t o compare the retention of radium by these people with that of the people studied by HURSHand GATES(6). The masiniuni permissible level recommended for RaZz6in drinking water is 40 p p c per liter ( 1 1 ) ; this is only 5 t o 10 times greater than the concentration of Razz6in some of the deep wells listed in Table 11. Radium i n Frankfurt and Pittsburgh Waters. Analyses were performed on water from two cities for which the radium concentration in the tap water had been reported to be high. KREBSreported 300 to 500 ,u,uc Ra per liter for tap water a t Frankfurt am Main, Germany, in 1939 (8). However, we found less than 0 . 3 ppc Ra per liter for the 12 sources of Frankfurt's drinking water now available and for the Main River (Table 111). One Frankfurt well was no longer in operation a t the time our samples were obtained. Check analyses on the Goldstein and Praunheim I1 samples using a radon emanation method showed 0.2 & 1 ,u,uc Ra per liter for each. I n the other case, BOTSETreported 143 p,uc Raper liter in a sample of tap water a t Pittsburgh, Pennsylvania (2). Many anaIyses of Pittsburgh water taken directly from reservoirs and rivers Argonne National Laboratory, but the largest concentration of Ra they found was 0.12 ppc per liter (13). Razz6 i n Water from Frankfurt am Main, Germany Samples collected 14/12/.50. - 2 Samples collected

Page 9

The cause of the above discrepancies is not known. In each case, the earlier results were mentioned as blank runs in reporting on work not concerned with the concentration of radium in drinking water. Possibly, the samples were contaminated at the water supply by an industrial establishment or during the analyses. KREBSreported his values on the basis of direct evaporation of water and counting of the residue, and BOTSETisolated Ra by precipitating BaSO,, metathesizing with carbonate, dissolving the BaCO, with HCl, and sealing the solution in an Erlenmeyer flask for a few days for radon to grow in for gamma-ray measurement. Thus, the activity measured by the former procedure could have been caused by many radioactive nuclides, and ThX would have been included in Botset's procedure. However, direct evaporation and counting of a sample of Pittsburgh tap water (collected 2/1,'52) in this laboratory showed less than 1 p p c of total alpha activity per liter one day after sampling, and approximately 1 p,uc of alpha activity per liter was found by direct evaporation …