i

Techniques of Water-Resources Investigations of the United States Geological Survey

Chapter A5 METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES IN WATER AND FLUVIAL SEDIMENTS

By L. L. Thatcher and V. J. Janzer, U.S. Geological Survey, and K. W. Edwards, Colorado School of Mines

Book 5 LABORATORY ANALYSIS UNiltED STATES DEPARTMENT OF THE INTERIOR

THOlMAS S. KLEPPE, Secretary

GEOLOGICAL SURVEY

V. E. McKelvey, Directsr

UNITED SATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1977,

For sale by the Superintendent of Documents, US. Government Printing Office Washington, D.C. 20402 Stock Number 024-001-02928-6 PREFACE This series of manuals on techniques describes methods used by the Geological Survey for planning and executing water-resources investiga- tions. The material is grouped under major subject headings called books and is further subdivided into sections and chapters. Book 5 is on labora- tory analysis, Section A is on water. The unit of publication, the chapter, is limited to a narrow field of subject matter. ‘‘Metho’dsfor determination of radioactive substances in water and fluvial sediments” is the fifth chap- ter to be published under Section A of Book 5. The chapter number includes the letter of the section. The looseleaf format of this methods manual is designed to permit flexibility in revision and publication. Supplements, to be prepared as the need arises, will be issued to purchasers at no charge as they become available.

111 i

i CONTENTS

Page Pase Preface ______--__-_--_---111 Gross alpha and beta radioactivity, dissolved Abstract ______--___-----1 and suspended. Residue method (R-1120- Introduction ______-_ -______1 76) ______-______29 Purpose ______-__---1 Reference ______------32 Organization ______-____ -- - _-_ _------1 -210, dissolved. Chemical separation and Nuclear data ______--- 2 precipitation method (R-1130-76) ------33 Units, symbols, and abbreviations ___--- 2 References ______------37 Sources of radioactivity in water ______2 , dissolved, as radium-226. Precipita- Natural radioactivity ______2 tion method (R-1140-76) ______39 Artificial radioactivity ______--- 3 Reference ______41 Permissible concentrations of radioactiv- Radium-226, dissolved. emanation ity in effluents to unrestricted areas ___ 3 method (R-1141-76) ...... 43 Radiological safety ______-- 4 References ______49 Geochemistry of radioactivity in water __ 4 Radium-228, dissolved. Determination by sepa- -14 ...... 4 ration and counting of -228 (R- Cesium137 and cesium-134 ______5 1142-76) ______------51 Lead-210 ______------5 Reference ______54 Radium ______------6 Radioruthenium, dissolved, as -106. Ruthenium-106 and ruthenium-103 - 7 Distillation method (R-1150-76) _-_-_-_-- 55 -90 ______- 7 References ______58 Tritium (-3) 7 - ______- Strontium-90, dissolved. Chemical separation 8 ______------and precipitation method (R-1160-76) 59 Collection and treatment of samples 9 __------References ______62 Calculations of radionuclide concentrations -- 11 Glossary ______-_--__---__-14 Tritium. Liquid scintillation method, Denver Selected references ______- 14 lab (R-1171-76) ______- 63 Principles of radioactivity, nuclear instru- Reference ______66 mentation ______------14 Tritium. Liquid scintillation method, Reston Compilations of data on radioactivity and lab (R-1173-76) ______67 radiochemistry ______-- ___ ------14 References ______71 Radioactivity in the environment ______15 Tritium. Electrolytic enrichment-liquid scin- Radioisotope methods in hydrology ______15 tillation method, Denver lab (R-1172-76) - 73 Radiochemical analytical methods ______15 Refepences ______-______78 Radioactivity regulations and safety ____ 15 Tritium. Electrolytic enrichmenGliquid scin- References ______-_-____------16 tillation method, Reston lab (R-1174-76) _ 79 Carbon-14, dissolved, apparent age. Liquid References ______81 scintillation method, Denver lab (R-1100- Uranium, dissolved. Fluorometric method- 76) ____r______~_~_~~~~~~~~~-17 direct (R-1180-76) 83 References 22 ...... ______References 88 Cesium-137 and cesium-134, dissolved. Inor- ______--__- ganic ion-exchange method-gamma count- Uranium, dissolved. Fluorometric method4x- ing (R-1110-76) ______----23 traction procedure (R-1181-76) ______89 References ______25 References ______- 92 Radiocesium, dissolved, as cesium-137. Inor- Uranium, dissolved, isotopic ratios. Alpha ganic ion-exchange method-beta counting spectrometry-chemical separation (R- (R-1111-76) ______------27 1182-76) ______-_---_------93 Reference ______------28 Reference ______95

V VI CONTENTS FIGURES

Page 1. In-growth and decay of a daughter nuclide, significant time intervals --__ 13 2. Apparatus for collection of carbonates from a water sample ______-18 3. Vacuum line for preparation of acetylene and conversion to benzene ______19 4. Growth of -210 from pure lead-210 source ______- 36 5. Radon deemanation train and bubbler ______------44 6. Radon scintillation cell and housing ______~~~_~~~~~-~~-45 7. Apparatus for the distillation of ruthenium tetroxide ______-__ 56 8. Ostlund electrolysis cell ______--_-__-_-_--___-_-____---75 9. Stevens apparatus for fusion and mixing of sample and flux in uranium determination ______~_~______~_____~~__~~~~-~~~84 10. dish for use in Stevens apparatus ______-----_-85 11. Uranium caIibration curve ______---86

TABLES

Page 1. Radon fraction (e-") remaining after for specified times - 48 2. Recommended sample volumes, minimum, and reduced volumes for isotopic uranium analysis ______------94 METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES IN WATER AND FLUVIAL SEDIMENTS

By L. 1. Thatcher and V. J. Janzer, U.S. Geological Survey, and K. W. Edwards, Colorado School of Mines

Abstract referenced so that an experienced radio- Analytical methods for the determination of some chemist could set up the analytical method of the more important components of fission or neu- with reasonable assurance of success. Two tron adivation product radioactivity and of naturd exceptions are :the determination of tritium radioactivity found in water are reported. The re- by gas counting and the determination of port for each analytical method includes conditions for application of the method, a summary of the carbon-14.Because of the complexity of the method, interferences, required apparatus ’and reag- equipment and the extreme importance of ents, analytical procedures, calculations, reporting certain critical details in both the instru- of results, and estimation of precision. The fission mentation and operating procedure, any ab product considered are cesium-137, stnon- tempt to convey the fully detailed analytical tium-90, and ruthenium-106. The natural radioele- for ments and isotopes considered are uranium, lead-210, procedure direct duplication has a low radium-226, radium-228, tritium, and carbon-14. A probability fo’rsuccess. There is no substi- gross radioactivity survey method and a uranium tute for actual operating experience in labor- j ratio method are given. When two analytical atories equipped for the gas-counting deter- methods are in routine use for an individual isotope, mination of tritium and the determination of both methods are reported with identification of the specific areas of application of each. Techniques for carbon-14. For the above reasons, the de- the collection and preservation of water samples to scriptions of these two analytical procedures be analyzed for radioactivity are discussed. concentrate on the principles and major op- erating conditions involved. In several analytical methods, reagents or Introduction equipment are cited by proprietary name. This is due to inadequate information on Purpose chemical composition on which to base a chemical name, or to special requirements This manual describes the analytical meth- known to be met by the cited reagent or ods used by the U.S. Geological Survey for equipment. In every case equivalent products the collection and analysis of water samples that meet requirements may be substituted. for radioactive substancw. The analytical No endorsement is intended. methods are intended for the radiochemist who applies his expertise to the analysis of water. Adequate grounding in the principles Organization and practice of radiochemistry is assumed. Each determination includes a section on Therefore, such subjects as nuclear instru- “Application,” “Summary of the method,” mentation, statistics, and radiation charac- “Interferences,” “Apparatus,” “Reagents,” teristics are not discussed. References are “Procedure,” “Calculations,” “Report” (of given to several excellent textbooks available results), and “Precision.” The “Calculation” on these subjects. section under each determination differs Generally, each analytical method is de- slightly from the practice in chapter A1 in scribed in sufficient detail and is adequately that reference is made to a general equation

1 2 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS when possible. Radioactivity calculations for ml ----milliliter i different nuclides follow similar principles gal ----gallon (3.785 liters) which can be summarized in a general equa- m -----meter tion. The general equation, with its modifica- cm -- --centimeter tions to fit particular requirements, is con- mm ---millimeter sidered in a following section. in. ----inch (2.54 cm) R -----foot (30.48 cm)

Nuclear data g -----gram Data on half-lives were taken from the mg ----milligram “Chart of the Nuclides” (Holden and Walk- M -----mdadty of a solution er, 1969). Data on;decay schemes and ener- N -----normaJity of a solution gies of nuclear radiation were taken from meq ---milliequivalent “Table of Isotopes” (Lederer and others, psi ----poundsper square inch 1968). The “Chart of the Nuclides” was re- produced in “Radiological Health Handbook” Sources of radioactivity in water (U.S.Dept. of Health, Education, and Wel- fare, 1970), and data on the more importank Radioactivity in water may be of natural nuclides listed in Lederer, Hollander, and or artificial origin. The principal natural Perlman (1968) were also reproduced in the proceslses that bring radioactivity into water HEW publication. are the weathering of rocks containing radio- active minerals and fallout of cosmic-ray- produced nuclides. The major sourceis of arti- Unit, symbols, and abbreviations ficial radioactivity are the nuclear power in- Terms that are generally used throughout dustry, nuclear weapons testing, and the the text are listed below. Terms that are used peaceful applications of nuclear materials infrequently, usually in conneckion with one and devices. analytical method, are defined at the first us- age in the text. Terms that are used in calcu- Natural radioactivity lation of data are defined under the general The principal radionuclides introduced equation in the “Calculations” sections. naturally into surface and groand waters are Ci -----curie (3.7 x 10’’ disintegrations per second) uranium, radium-226, radium-228, radon, pCi ----microcurie (3.7 x 10‘ disintegrations per -40, tritium, and carbon-14. All second) but the last two derive from radioactive pCi ---picmurk (3.7 x lo-’ disintegrations per sec- ond) minerals. Radioactive elements including pCi/l --microcuries per litm uranium, , and actinium and radio- pCi/l --picocuries per liter active daughters resulting from these de- cay series are important primarily for cpm ---counts per minute dpm ---disintegrations per minute reasons of health and as potential energy sources. These three natural decay series are keV ,--thousand electron volts headed respectively by uranium-238 (half- MeV --million electron volts life 4.51 x lo9 yr) , thorium-232 (half-life 1.41 1O1O yr) , and uranium-235 (half-life d ---,--day x hr --,--hour 7.13 x los yr) . In areas of the world where min ---minute radioactive minerals are particularly abun- sec ----second dant as in the Joachimsthal region of Czecho- yr ---- slovakia, the Minas Gerais region of Brazil, and the Colorado Plateau of the United e __-___2.718.. .. ., base of the natural logarithms In -----logarithm of any number N, to the base e States, radioactivity in some waters may log ----logarithm of any number N, to the base 10 greatly exceed the average concentrations METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 3

present in most continental waters. Dissolved by 1980, several million gallons of high-level I natural uranium is the major with radioactivity oa the order of 10 constituent in most of these waters. billion curies will be in storage as a result of Tritium and carbon-14are produced by the nuclear power produckion (Hogerton, 1963, interaction of cosmic-ray neutrons with p. 448). in the upper atmosphere. The triti- A relatively smaller amount of radioactivi- um is eventually rained out as tritiated ty may leak to the environment as a result of water, and the radiocarbon is incorporated daily operation of a nuclear powerplant be- into atmospheric carbon dioxide. The princi- cause of neutron activation of salts in the pal reservoir for both radionuclides is ulti- coolant water, neutron activation of dis- mately the ocean. Both radionuclides are also solved corrosion products, and possible re- produced by thermonuclear weapons testing. lease of fission products by a defective fuel In 1963, the year when radioactive fallout element. Although every nuclear powerplant reached its maximum, the atmospheric con- has built-in safeguards against release of centration of thermonuclear tritium exceeded radioactivity, the possibility of accidental that of natural tritium by approximately 3 leakage must always be considered. orders of magnitude. The additiond carbon- In addition to tritium and carbon-14, 14 of thermonuclear origin wm much lower aboveground nuclear weapons testing re- and approximately equalled the carbon-14 leases strontium-90, radiocesium isotopes, naturally in the atmosphere. Radioisotope -131,and other nuclides to the environ- concentrations in fallout have diminished ment. The fraction of these radionuclides rapidly since 1963 as a source of radioiso- that fall out or rain out into water bodies or topes in water. Tritium is also a fission prod- watersheds constitute a significant source of uct, and by 1970 the nuclear power industry water contamination. had probably become the largest source of Peaceful applications of nuclear explo- l tritium (Jacobs, 1968). sives, such as nuclear gas stimulation and nuclear mining, present a possible source of Artificial radioactivity locally intense contamination of ground Nuclear waste disposal is the principal water. If venting occurs, the possibility of possible source for the leakage of artificial contamination of surface water by fallout radionuclides (fission products and activa- also exists. tion products) into water. Most of the waste Nuclear research laboratories, hospitals, is derived from the reprocessing of nuclear and the very limited number of industries fuel. Reprocessing is required at intervals to that use radioactivity constitute a relatively remove neutron-absorbing fission products minor source of possible radionuclide leak- and to recover the uranium and . age, but may be of local significance. Until reprocessing, 99.9 percent of the fission Measureable radioactive material may de- and activation products produced remain rive from (sources not normally considered locked inside the fuel element. The fuel ele- radioactive. Examples are fly ash from the ment is dissolved in acid, and chemical sepa- combustion of fossil fuel (coal may contain rations of the highly radioactive wastes are uranium and radium) and the ceramics in- carried out. The final waste consists of a low- dustry (uranium salts are used in the prepa- level solution, which may be sent to seepage ration of some glazes). ponds, and a high-level solution or “hot” solu- tion that must be stored for many to Permissible concentrations of allow radioactive decay. Strontium-90 (half- life 28.9 yr), cesium-137 (half-life 30.2 yr), radioactivity in effluents to iodine-129 (half-life 1.6 X lo7 yr) and plu- unrestricted areas tonium-239 (half-life 24,390 yr) are major Current values are tabulated in Part 20, radioisotopes of concern. It is estimated that “Standards for Protection Against Radia- 4 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS tion,” published and updated periodically by Energy Agency includes reports of particu- i the U.S. Nuclear Regulatory Commission lar value to laboratory users of radioactivity. (1976). The Commission is the source of the “Safe Handling of Radioisotopes” (1962) permissible cancentrations of radioactivity and the “Health Physics Addendum” (1960) in effluent (PCRE) values reported for the are cited. Precautions against contamination nuclides in this manual. The significance of and procedures for decontamination are de- the values and the basis for their computa- tailed in the National Bureau of Standards tion is described in National Bureau of publication “Control and Removal of Radio- Standards (NBS) Handbook 52 (1953) and active Contaminakion in Laboratories” NBS Handbook 69 (1959). The U.S. Public (1951). Procedures to be followed in event of Health Service hm published recommended an accident involving radioactivity are de- maximum concentrations for radium-226, scribed in “Medical &pets of Radiation strontium-90, and gross beta activity in accidents” (Saenger, 1963). drinking water (Drinking Water Standards, 1962). Updated standards published in “Wa- ter Quality Criteria, 1972” (1973) are quoted Geochemistry of radioactivity in addition, when applicable. in water Additional drinking-water regulations were published in the Federal Register, vol- Dissolved and particulate radioactivity in ume 41, No. 133, July 9, 1976, as a supple- water is controlled by the same mechanisms ment to Title 40,Code of Federal Regulations that affect other trace and macro constituents (CFR), Part 141. The National Interim Pri- in the geohydrologic environment. Radioac- mary Drinking Water Regulations (Part tive disintegration of an atom by alpha and 141) have been published (1975), but were beta decay results in the formation of an in the process of revision at the time this atom of a new element, frequently in an ex- paper was written. cited state. Gamma emission results from I such an artom in the excited state going to a lower energy state. The geochemical behavior Radiolog ica I safety of a daughter element may be grossly differ- The radiochemist or chemist who uses the ent from that of the radioactive parent, al- methods in this manual should have an ade- though its occurrence, distribution, and quate training in radiological health and transport may be governed by the parent. safety practice in the laboratory. Such train- ing is a requirement for obtaining the U.S. Carbon-14 Nuclear Regulatory Commission license to Carbon-14 is the radioactive isotope of use radionuclides. Analysis of even environ- carbon with a half-life of 5,730 years. The mental-level samples will generally require older half-life value of 5,568 yr is generally the use of various radioactive calibration used for the calculation of ages in order to planchets or standardized solutions, some of maintain consistency with carbon-14dates in which are hazardous if not properly used. the older literature. Ages based on the older Discussions of various aspects of radiological half-life value are converted to the basis of safety are given in the following suggested the newer half-life by multiplying by 1.03. references. An introduction to the subject is Carbon-14 decays by the emission of a beta found in the chapter on the radiological particle with the low maximum energy of 156 laboratory in “Guide for Safety in the Chemi- keV. cal Labo’ratory” (M.C.A., 1954). A compre- Neutrons produced by primary cosmic hensive exposition of radiological health radiation interact with stratospheric nitro- physics is found in “Principles of Radiation gen to produce carbon-14 and hydrogen : Protection’’ (Morgan and Turner, 1971). The safety series of the International Atomic I;N +pl:C + :H. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 5 , The rate of production in the atmosphere in the Southern Hemisphere as a result of is estimated at approximately 2.4 atoms per nuclear weapons testing. second per square centimeter of the Earth’s Either the beta OF gamma radiation associ- surface (Libby, 1955). Carbon-14 is also a ated with the cesium isotopes may be used for product of thermonuclear weapons testing. their dhction. Detection by beta radiation This source had approximately doubled the is more sensitive but less specific. The use of atmospheric concentration of carbon-14 and gamma radiation permits distinguishing be- had increased the concentration in surface tween the isotopes by means ob energy dis- ocean water by about 20 percent by the late crimination since the energy peak for cesi- 1960’s (Nydal, 1966). Possible contaminac um-137 (-137m) is at 662 key and the tion of ground waters by thermonuclear car- principal peaks for cesium-134 are at 605 bon-14 imposes a severe limitation to the keV and 796 keV. Preconcentration of the application of carbon-14dating. radiocesium is used before counting, in ei- Carbon-14produced by cosmic radiation is ther case, to permit detection at low environ- oxidized to carbon dioxide and is transported mental levels. Four to 20 liters 02 sample to the lower atmosphere by mixing processes have been used with the gamma-counting where it enters the biological cycle. Some technique depending on sensitivity required. radioactive carbonates enter the hydrological A few hundred milliliters are usually ade- cycle and provide the basis for carbon-14dat- quate for the beta-counting technique. ing of older ground waters. The specific ac- tivity of cosmic-ray produced carbon-14 in Lecld-2 1 0 the atmosphere, surface waters, and all liv- Lead-210 originates from the decay of ing matter was determined by Libby (1955) radon-222.The isotope is a beta emitter with to be 16 dpm per gram of carbon, but is now half-life of 22 years. Lead 210 formed under- considered to be lower. Suess (1965) deter- ground is probably trapped on exchange sites i mined 14 dpm per gram of carbon. of clay minerals or other reactive material The specific activity of carbon-14 in car- and presumably has very limited migration. bonaceous material cut off from contact with The lead-210formed by decay of atmoepheric the atmosphere,such as the carbonate species in ground water, decreases at a rate con- radon enters the hydrologic cycle principally trolled by the carbon-14 half-life. The car- through precipitation. A smaller part is re- bon-14 to carbon-12 ratio is also affected by moved from the atmosphere as dry fallout. exchange of carbonates between the water Most of the lead-210 falls on the Oceans and aquifer, biochemical de&, and possible where it finds use as a tracer far the investi- reactions with silicates. Compensation for gation of vertical mixing. Some lead-210 en- exchange effects has been attempted through ters terrestrial surface waters and the the carbon-13 to carbon-12 ratio (Pearson, permanent snow fields. The half-life is con- 1965). venient for dating more recent snow deposits, and the natural level of lead-210 has not been Cesium-137 and cesium-134 upset by nuclear wapons testing as is the Eleven cesium isotopes are fission prod- case with tritium. Estimation of the lead-210 ucts, but usually only cesium-137 is signifi- input through precipitation, a requirement cant to water quality. Cesium-134 is pro- for dating of snow and ice, must be based on duced in fission by the neutron activation of actual measurements of the lead isotope since cesium-133,a fission product. The half-life of calculation on the basis of equilibrium with cesium-137 is 30 yr and that of cesium-134is prevailing radon concentrations is not ac- 2.7 yr. Cesium-137 has been deposited curate. This was shown by Patterson and throughout the world, with much higher con- Lwkhart (1964) who made a latitudinal sur- centrations in the Northern Hemisphere than vey of lead-210 in ground-level air. 6 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS The lead-210 content of precipitation ap- and the specific activity of thorium-232is ap- pears to vary greatly with individual storms, proximately one-third that of uranium-238, probably depending on the trajectory of the the world activity inventories of radium-226 air mass. Rama and Goldberg (1961) re- and radium-228 should be roughly equal. ported 0.05 pCi/l for surface ocean water World activity inventory for radium-224 and 0.13 to 6.7 pCi/l for Colorado River wa- should also be equal to that of radium-228 ter. Holtzmann (1964), using the more sensi- since the two are in radioactive equilibrium. tive -210 alpha-counting technique, The local relative abundances of the 226 and found an average 09 0.127 pCi/l of lead-210 228 radium isotopes may vary greatly, how- in untreated Illinois surface waters and 0.051 ever, as a function of the local uranium to pCi/l in well waters. As expected, the sorp- thorium ratios. Also there may be extreme tion of lead by soil reduces the concentration local variations in the ratio of radium-228 to in ground water. Holtzmann found that lead- radium-224 because the latter is produced 210 concentrations in potable waters of Illi- from the former through an actinium and a nois were generally much below radium-226 thorium isotope. Because the geochemistry of concentrations. actinium and thorium are significantly differ- ent from that of radium, there is great op- Radium portunity for local disequilibria. Radium is a radioactive member of the Radium is found in waters from most geo- alkaline-earth family that is widely dissemi- logic terranes because of the wide distribu- nated throughont the crust of the Earth. tion of the parent elements in nature. Con- Four are members of the centrations of radium-226 in freshwater three natural radioactive series. The isotopes usually are less than 1-2 pCi/l. Concentra- with their natural series and decay data are tions of radium-226 in the Cambrian and listed below. Ordovician limestones of North Central United States often exceed 3 pCi/l, and in Decay Isotope Series Half-life mode certain aquifers of the Colorado Plateau, the Radium-223 ____ Actinium _____ 11.43 d a concentration may be much higher (Scott and Radium-224____ Thorium _____ 3.64 d a Barker, 1959). Water that leaches waste Radium-226____ Uranium _____ 1,602 yr a piles from uranium mining and milling op- Radium-228 ____ Thorium _____ 5.75yr B erations may contain radium at much higher The concentrations of the radium isotopes levels. in geologic and hydrologic materials vary While most of the radium investigations greatly in nature depending on the uranium have centered on the 226 isotope, work in the and thorium concentrations in the source and U.S. Geological Survey has shown the im- the geochemical history ob the material. portance of radium-228.Johnson (1971) re- Chemically, radium is analogous to bari- ported that concentration of radium-228 in um;the carbonates, sulfates, and chromates several streams of the Front Range near are insoluble, while the chlorides, nitrates, Denver exceeded the concentration of radi- and hydroxides are soluble in water. The dis- um-226. This agrees with the twice normal tribution of radium is governed more by the abundance ratio of thorium to uranium for distribution of uranium and thorium, how- the area. Krause (1959) reported relatively ever, than by the geochemistry of radium. high concentrations of radium-228 in wells Radium-226 and radium-228 are the most tapping deep ground-wateraquifers in Iowa, important isotopes of radium found in water Wisconsin, Illinois, and Missouri. because of their Ionger half-lives, health sig- All radium ilsotopes are hazardous because nificance, and as geochemical indicators of of the bone-seeking properties of the ele- uranium and thorium respectively. On the ment. Concentrations in the bone can lead to basis that the world abundance of thorium is malignancies. The US.Public Health Service approximately three times that of uranium has recommended 3 pCi/l as the upper limit METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 7 for radium-226 in water for public consump- 1150-76 are reported as ruthenium-106,al- tion. The U.S. Nuclear Regulatory Commis- though ruthenium-103 may also be present sion (NRC) (1976) gave the following per- in the sample. After 8 months decay (6 half- missible concentrations of radioactivity in lives) , the ruthenium-103 concentration is effluent (PCRE)values for radium (soluble) usually insignificant. in waste solutions that may be released to water bodies accessible by the public : Strontium30 Radium-223 ______700 pCi/l Nuclear fission produces two important Radium-224 ______2,000 pCi/l : strontium-90 (half- Radium-226 ______30 pCi/l life 38.9 yr) and strontium-89 (half-life 50.8 Radium228 ______30 pCi/l d). Although the latter nuclide has initially The “Water Quality Criteria, 1972” (EPA, greater activity, the longer half-life of stron- 1973) recommended a radium-226 intake tium-90 makes it more significant to world- limit of 0.5 pCi/d. Assuming a 2 l/d con- wide environmental pollution. Except for sumption rate, this is equivalent to a 0.25 short periods following atmospheric nuclear pCi/l concentration limit. testing, strontium80 is the predominant Three analytical methods for radium iso- radioisotope of this element on the Earth’s topes are reported, each serving a different surface. This nuclide is now widely dis- purpose. A determination of the gross alpha tributed in man’s environment from strato- radioactivity of radium is the simplest pro- spheric and tropospheric fallout. Higher con- cedure and is satisfactory where identifica- centrations exiist in the Northern Hemi- tion of individual alphsGemitting isotopes is sphere. It is often detected in soils, foods, not required. A second procedure is specific water, and biological materials. The “Water for radium-226 and a third for radium-228. Quality Criteria, 1972” (EPA,1973) recom- mended limit on strontium-90 intake in Ruthenium-106 and ruthenium-103 water used for public supply is 5 pCi/d. Ruthenium-106 and ruthenium-103 are the Assuming a 2 l/d consumption rate, this is important ruthenium isotopes produced in equivalent to a 2.5 pCi/l concentration limit. nuclear fission. They may be present in pre- cipitation and surface waters after atmos- Tritium () pheric nuclear testing. The short half-life of Tritium is the radioisotope of hydrogen ruthenium-103 (39.8 d) essentially rules out with atomic weight of 3. It decays by pure its presence in ground water. Ruthenium-106 beta emission with half-life of 12.33 yr. The (half-life 368 d) may be found in ground beta particles have an average energy of 5.7 water in the immediate vicinity of under- keV and maximum energy of 18.6 keV. Triti- ground nuclear tests. Both isotopes are beta um is formed in the upper atmosphere by emitters with maximum energy of 0.0392 cosmic-ray spallation and by the interaction MeV for ruthenium-106 and maximum ener- of fast neutrons with nitrogen : gy of 0.70 MeV (3 percent) and 0.22 MeV ‘:N +;n+;H +.;C. (97 percent) for ruthenium-103.Ruthenium- 106 is determined by counting the daughter, The natural production rate of tritium is -106,with which it is in secular equi- on the order of 30 atoms per square centime- librium. The more energetic beta particles of ter of the Earth’s surface per minute. (Kauf- rhodium-106 (3.53, 3.1, 2.4, and 2.0 MeV,) man and Libby, 1954). are more easily detected. Tritium ias also produced by thermonuclear The NRC-PCRE value for ruthenium-106 weapons testing. The first tritium from this in effluents released to uncontrolled areas is source was detected in 1952, and by 1954 the 1 X pCi/ml (10,000 pCi/l). Ruthenium thermonuclear tritium was substantially concentrations determined by procedure R- greater than the natural tritium. (Begemann 8 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS and Libby, 1957). Thermonuclear tritium tritium inventory is substantial. Sources of reached a peak in 1963 when the concentra- tritium have been reviewed by Jacobs tion in the Northern Hemisphere exceeded (1968). the natural level by approximately 3 orders Determination of tritium radioactivity is of magnitude. complicated by the very low energy of the Tritium is principally of interest to the beta radiation which necessitates mixing the hydrologist as a water-dating tool and as a tritium intimately with the counting medi- tracer, introduced either naturally or arti- um. Gas phase counting is carried out by ficially, for investigating ground-water hy- introducing the tritium as a gas (HT) into a drodynamim in areas of relatively rapid flow. proportional counter containing the proper The tritium dating of ground water that en- pressure of hydrocarbon gas to assure opera- tered as recharge before 1952 is based on tion in the propo;rtional region. Liquid scin- radioactive decay of tritium and the ‘‘pre- tillation counting of tritium is carried out by bomb” concentration of approximately 8 tri- mixing the tritiated water sample with an tium units (Tu), (1 Tu=l T atom/1018 H organic scintillator suspended or dissolved in atoms) determined by Kaufmann and Libby an organic medium compatible with water. (1954). The efficiency of gas phase counting may be The dating of ground water originating as very high, 60-80 percent. The efficiency of recharge after 1954 is based on the correla- liquid scintillation counting is on the order tion of tritium concentrations in the ground of 20-25 percent. water with known fallout peaks. The applica- Since the decay rate for 1 Tu concentration tions of tritium in hydrology are reviewed by in water is only 0.007 dpm/ml and the back- Thatcher (1969). ground count (gas counting) is on the order Tritium is also of interest in public health of 2 cpm,it is obviouls that enrichment of the inasmuch as concentrations much exceeding tritium in the water sample is required when the natural level may be released to the en- low concentrations of tritium are to be de- vironment in the course of tritium tracer termined. Enrichment is carried out by elec- experiments, nuclear po,wer production, trolysis using essentially the same process as weapons testing, nuclear waste disposal, and is used for the preparation of heavy water. “Plowshare” activities. “Water Quality Cri- Passage of electric current through water re- teria, 1972” (EPA,1973) notes that a tenta- sults in the liberation of the light hydrogen tive limit of 3,000 pCi/l of tritium has been isotope (protium) at a faster rate than the prepared for the revised edition of “Drinking heavy isotopes (deuterium,tritium), thus en- Water Standards.” The NRC (1975) PCRE riching the latter in the residual water. value for tritium in effluents released to un- controlled areas is 3 pCi/l. The liquid-scintil- Uranium lation counting technique provides all the Uranium is widely disseminated in the sensitivity required for monitoring tritium lithosphere, and most natural waters contain concentrations at the levels significant to detectable concenlxatiom of this element. public health. The average Concentration in the ocean is Although tritium is not a major fission about 3 pg/l (Rona and others, 1956). The product, there is a significant production of uranium content of ground and surface wa- tritium in nuclear power reactors. The great- ters varies greatly, from less than 0.1 pg/l to er part of the tritium remains enclosed in several milligrams per liter (mg/l). In most the fuel element until the latter is dissolved natural waters the concentration is less than for reprocessing. The production of tritium 10 pg/l. The limit on uranyl ion (U02S2)in in reactors depends greatly on the reactor public supplies (Water Quality Criteria, type, and estimates of total production are 1972) is 5 mg/l. subject to many uncertainties. It seems clear, Dissolved uranium in natural waters exists however, that the reactor contribution to the principally as uranyl ion (U02f2)which may METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 9 be complexed with carbonate. While uranium 234 in water (Chalov and others, 1964;Cher- in surface waters is almost entirely hexa- dyntsev and others, 1961; Thurber 1962). valent, there may be an appreciable percent- Depletion of the 234 isotope is less common. age of tetravalent uranium under reducing Isotopic dating of waters, correlation of dis- conditions found in some ground waters. The equilibria with rapid leaching (important in latter is predominantly complexed and is pollution studies), and the geochemistry of highly insoluble in basic solution. uranium transport are examples of the possi- Uranium in natural water is important in ble application of uranium isotope disequi- geochemical prospecting and as an indicator libria studies. of pollution from mining operations. It is un- While the usual objective of the analysis likely that uranium could reach concentra- is determination of the uranium-234 to 238 tions significant to health, although a high activity ratio, it is also possible to determine concentration of uranium could be indicative the activity ratio of the 235 isotope to 238. of possible high levels of the much more The uranium-235 to 238 ratio is constant hazardous radium isotopes. (1: 137) in the absence of artificial depletion There are the three natural uranium or enrichment. Therefore, an increase of the isotopes : 235 isotope is reason to suspect the presence of material processed for nuclear fuel. Abundance Half-life ISotQPe (percent) (yr) U-238 (U series) --_------99.27 4.51 x 10’ Collection and treatment U-235 (Ac series) ______.72 7.1 x108 U-234 (U series) ______.006 2.47X1O5 of samples Uranium-234 is related to uranium-238, The principal requirements to be met in the parent of the uranium decay series, by sampling a water body for the determination the following decay sequence : of radioactive constituents are the same as for other constituents, that is, collection of Uz3s~Thz34(24.1 d),Pa234P (1.18 min),U234.P samples in an orderly sequence that provide a The total world radioactivities of the 234 representative analysis of the water body and 238 isotopes must be equal since the two both areally and temporally. Considerations are in secular equilibrium. Local physical and involved in setting up a sampling program chemical effects may result in local disequi- that yields the required representative inf or- libria which are of geochemical interest and mation are discussed in detail in an earlier have been given intensive study. Disequi- chapter (Brown and others, 1970) in this libria may result from physical or chemical series of Techniques of Water Resources In- mechanisms. The principal physical mechan- vestigations of the U.S. Geological Survey. ism is the energetic alpha recoil associated Reference is made to the earlier chapter for with the decay of uranium-238 to thorium- discussion of site selection, sampling fre- 234. This may rupture chemical bonds and quency, equipment,sample identification, and permit thorium to go into solution in ground other elements of an organized sample-collec- water where it demys to uranium-234. Al- tion program. Guidelines for the collection though thorium is considered to be sorbed by and field analysis of ground-water samples sediment, the presence of dissolved organic are given by Wood (1976). If analyses for matter or other complexing agents may tend the determinations described in this manual to stabilize it. Since both intermediates, are to be made by a U.S. Geological Survey thorium-234 and protoactinium-234, are Water Resources Division central laboratory, chemically differentfrom uranium, chemical the appropriate section chief should be con- differentiation may occur. Difference in oxi- tacted. Specific information can then be ob- dation state between the 234 and 238 isotopes tained for collecting the samples, obtaining can have an additional effect. These effects the necessary sampling supplies, and so frequently lead to enrichment of uranium- forth. 10 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS The overriding problem in sampling for reported in the literature are based on dis- radioactivity is preservation of the extremely tilled-water solution, and thus are not entire- dilute concentration of radionuclides (usually ly relevant. in the nanogram- and picogram-per-liter Similar tests with uranium showed no sig- range) in their original conditions until the nificant lolss under any condition. Hexavalent analysis can be performed. While preserva- uranium is stable in natural water exposed to tion of the original condition of the sample is the atmosphere, particularly when bicarbon- a general problem in water analysis, the ate is present. Earlier tests by Janzer (un- extreme dilution of the radionuclides greatly pub. data) are in agreement with the above intensifies the problem. Although extensive findings. Surface-water samples were col- research has been given to the chemistry of lected in polyethylene bottles under three extremely dilute solutions, the practical re- conditions : no treatment, filtered and acidi- sults in terms of preservation of the sample fied at time of collection, and filtered at time have not been completely effective. Indeed, of collection but not acidified. The coacentra- conclusions from different investigations are tions of radium, uranium, and gross alpha- often at variance, and indicate that the essen- beta activity determineld repetitively in all tial phenomena are not completely under- three samples agreed closely and did not sig- stood, nificantly change with time. Early in the history of radiochemistry the It is generally recommended that acid term “radiocolloid” came into use as a result should be added to very dilute solutions of of apparently colloidal phenomena observed trace elements for the purpose of minimizing in radioisotope solutions (dialytic separation, sorption loss of trace elements from the solu- nonionic migration behavior in the electric tion while it is in contact with a sample con- field, and so forth). Although radioactivity tainer. ThPs is the practice recommended produces electrically charged centers (which herein, but it should be noted that the experi- could attract other ions to form aggrega- mental evidence is not at all conclusive. There \ tions), it is unlikely that the phenomena in appears to be significant evidence from both very dilute solutions of radioisotopes are radioactive and nonradioactive work with greatly different from phenomena in very trace elements that the individual chemistry dilute solutions of nonradioactive ions. Starik of the elements must be considered, and op- (1959) has exhaustively reviewed the earlier timum preservation techniques for each ele- work and finds both colloidal and noacolloidal ment, or for chemically related groups of behavior depending on conditions. elements, are required. Starik (1959) re- Starik also investigated the effect of filtra- ported, fo’rexample, that polonium is sorbed tion and found that radioisotopes were re- most strongly (on glass) at pH 4.5, and while tained in varying ratios by different filter the sorption is much less below pH 4, it is media depending upon the pH and other reduced to the minimum above pH 8. He also chemical factors. Overall, both sorption loss reported minimum sorption of radiotantalum and filtration loss are functions of the con- at pH 10 with maximum sorption at pH 3.5, centration, pH, oxidation state, electrolyte and maximum sorption of -95 at pH composition, and presence of traces of col- 2. loidal material in the solution such as col- Ground-water samples are usually clear loidal silica as well as numerous other initially but may become turbid as hydrated factors. Many observations of “radiocolloids” and oxides form on exposure are apparently attributable to sorption of the to air. It is essential to prevent the formation radionuclide on traces of colloidal silica. of these precipitates because they can copre- Tests suggest that the sorption loss in cipitate radioactive elements. Precipitation natural-water samples, as against sorption of hydrous oxides is prevented by acidifying loss in very dilute distilled-water solutions, the sample after filtration. Sufficient reagent may be relatively small. Many of the tests grade hydrochloric or nitric acid should be METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 11 added to the filtered sample to obtain a pH of more acid. Test with pHpaper. Seal the 1 or less. cap with vinyl tape. A 4-liter sample is usually adequate for 3. At the time of collection, fill out appropri- radiochemical analysis of gross alpha, gross ate sample data form as completely as beta, uranium, and radium. Additional Sam- possible and include with the sample in ple may be required if other analyses or re- the manner prescribed by the analyzing runs are desired. Recommended sampling laboratory. When compiling the results procedures for surface and ground waters of the analysis, it is essential to have are described as follows : all the information possible pertaining to the conditions which existed at the 1. Collect and store the samples only in spe- time the sample was collected. cially washed and labeled 4-liter poly- 4. Box the samples in the cartons provided, ethylene sampling bottles provided by and ship as soon as possible after col- the radiochemical laboratory. The bot- lection. Be sure that return address is tles are cleaned by the following pro- on the shipping label. cedure: Wash with tap water; add 10 5. During winter months, it is essential to nitric acid, and refill tap water ml with ; mark the cartons “Water Sample. Keep allow to stand overnight; empty, and from freezing.” finally rinse several times with small amounts of distilled water ; and drain, When sampling a ground-water system for and allow to dry before capping and tritium or carbon-14, it is particularly im- storing. portant to select the sampling wells with care 2. Surface-water samples are collected to ob- so that a representative sample is obtained. tain a representative sample of well- The well should be properly sealed to mini- mixed water at a single point, generally mize surface-water contamination, and it near the center of the stream or river if should preferably be in constant use. A high- possible. Preferably only clay- or silt- !y productive well is preferred to a low-yield sized particles should be present in the well. The well should be thoroughly pumped sample if samples are to be analyzed before the sample is taken. A perfectly dry for suspended gross alpha and beta bottle or barrel is used. During and after the radioactivity. No filtration or acidifica- collection of the sample, minimize contact tion is usually required if dissolved and with the atmosphere which may contain car- suspended gross alpha and beta radio- bon-14 and tritium at concentrations ranging activity and dissolved radium or urani- from several-fold to several orders of magni- um are desired. (Other analyses may tude higher than the radionuclide in the require special handling, and the ap- water sample. propriate Central Laboratory section chief should be contacted for specific Calculations of radionuclide details.) Leave an air space of several centimeters to allow for volume concentrations changes with temperature. Seal the cap The metho,d used to calculate concentra- with vinyl tape. Ground-water samples tions of radionuclides for most of the deter- should be filtered through a 0.45-mi- minations in this manual may be expressed crometer membrane filter at the time of in the form of a general equation. Exceptions collection. Add sufficient reagent-grade are the calculations for uranium, uranium hydrochloric acid (preferred), or nitric isotope ratio, and carbon-14age. The general acid, to lower the pH to approximately method of calculation compares the activity 1. Minimum value of concentrated acid of a sample against the activity of a stand- that should be added to a 4-liter sample ard, and corrects for decay of the radionu- is 8 ml. Alkaline waters may require clide in the sample between time of collection 12 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS specified for the determination. E and time of analysis. If the recommended i practice of minimizing time delay between is usually determined by analyzing sampling and analysis is followed, the correc- standards in the same analytical tion for decay throngh this interval becomes procedure as for samples (a modi- negligible for many radionuclides. fied procedure is used in two meth- A more significant correction is that neces- ods described in this report), sitated by decay of the standasd between the f=fractioaal recovery of the nuclide in date of its certification by the National Bu- the Sample, reau of Standards (or other supplier) and x=decay constant of the nuclide deter- the date of its use to calibrate the analytical In 2 mined by: 7, method. This time interval may amount to 1 several years and is significant, relative to YZ the half-lives of several radionuclides. where The following general equation applies T,=ha.lf-life of the nuclide of in- when in-growth of daughter activity is not a terest in the appropriate factor, and when the half-life is long relative time units, and to the counting time. This is the usual t = elapsed time between collection of the situation. sample and count of radioactivity (in same time units as used for A). The co,untingefficiency factor E,is calcu- lated by the follo~winggeneral equatioa : where C = concentration of radionuclide. This is -_usually expressed in pCi/l. c=cs- (b,+b,).Average count rate of where! - \ the sample in counts per minute c,=average count rate of standard in (cpm) after co,rrectionfor back- cpm after correction fw back- ground and blank, ground and blank, where - d, = disintegration rate of standard a, =average gross sample count rate (dpm) , - 9 (c??m) fn = fractional chemical recovery of the b, =average blank count rate nuclide in the standard, and - (cpm), and b, = average background count rate t, =elapsed time between certification of the standard and the count, in (CR-4- Usually b, and x2are experimental- some units as the respective A. ly determined as a combined In the determinations of gross alpha and quantity. beta activity, cesium-137 and 134, tritium K = factor to convert disintegration rate (without electrolysis) , radium-228, radium in disintegrations per minute as radium-226,and radium-226,the chemical (dpm) to curies. The value of K recovery factors f and f, are not determined. for different concentration expres- The product Ef, is determined by counting sions (C) is: the standard and is substituted for Ef in C K equation 1. This procedure is valid when f mCi/l ______2.22~10~ and fn are equal and reproducible (within ex- pCi/l ______2.22~10~ perimental limits) and effectively cancel out. pCi/I ______2.22 The usual situation is that f and f, are very V = volume of sample in milliliters, close to unity. E= coumting efficiency for the nuclide Equations 1 and 2 apply when the sample under the counting conditions and standard are counted under the me METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 13

, conditions on the same detector. This is the I I I normal situation. In the determination of - radium-226the individual samples and stand- /---- ards are counted in individual alpha scintilla- /-- - 0* tion cells, each of which has its own counting /#e- efficiency (cell constant). Hence, in the de- termination of radium-226 an individual counting efficiency is determined for each cell. In the determination of lead-210, radium- 226, and radium-228,the radioactivity count is made on a relatively short-lived daughter of the nuclide determined. This necessitates the introduction of an “in-growth” or “build- up” factor, which is the fraction of the equi- librium concentration of daughter that had grown in at the time of separation from the parent nuclide. Since the daughter nuclides are relatively short lived with respect to the counting time, it is necessary to introduce a correction factor for decay during the count- ing interval. With radium-226 and radium- 228 it is also necessary to introduce a factor that corrects for decay of the daughter dur- ing an aging prior to counting. j Although a daughter nuclide is counted (in addition to the parent) in the determination or“ strontium-90, an in-growth factor is not used because the daughter is allowed to reach a constant level (99.5 percent of equilibri- um) before counting. The relationship of the three correction Figure 1.-In-growth and decay of a. daughter nuclide, factors to the time intervals involved in the significant time intervals. counting of a daughter nuclide is illustrated in figure 1. The figure shows growth and de- Y=e-Vz (decay of daughter between cay of the daughter with time and identifies separation from parent and begin- the significant time intervals. The general ning of co,unt), equation for use ingrown nuclides is with : V8 z= (decay of daughter during 1- e-V3 counting period), where where C = concentration of radionuclide. This A= decay constant of parent nuclide, is usually expressed in pCi/l. - t = elapsed time of parent between c=average count rate of sample in cpm collection of the sample and after correction for background separation of daughter, and blank, x1= decay constant of daughter nu- X = 1- e-% (in-growth of daughter), clide, 14 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS t, = in-growth time of daughter, Filterable solids. Those dissolved solids cap- 1 point A to B (fig.l), able of passing through a 0.45 micrometer t, = delay before counting, point B to membrane filter and dried to constant C. Separation of the daughter weight at 180°C. from the parent at time B,and Half-life. The time required for the decay of t, = time interval for counting of the a given quantity of a radioactive substance daughter, point C to D. to one-half its original mass; thus it is a K,V, f, and E are as defined for equa- measure of the rate of such processes. tion l. The efficiency calculation for ingrown Minimum detection level. The least amount nuclides is: or concentration that can be detected and quantified by a test method. (4) Nonfilterable solids. Those solids which are retained by a 0.45 micrometer membrane with symbols and units as defined in equa- filter and dried to constant weight at 103"- tions 2 and 3. 105°C. Because of the short counting time permit- Precision. The degree of agreement of re- ted by the higher concentration of the stand- peated measurements of the same property ard, the counting time is short relative to the expressed in terms of dispersion of test half-life. Hence the correction factor for results about the mean result obtained by decay during counting is eliminated. testing of a homogeneous sample (s) under The general equations are modified in ac- specific conditions. cordance with specific conditions prevailing Sorption. A general term for the processes of in the determination of individual radionu- absorption and adsorption. clides. For example, f and fR are eliminated when chemical separations are not used or when the chemical recovery facto,ris included Selected ref,erences i in the determination of overall efficiency. Decay terms are eliminated when the half- life of the nuclide permits. Principles of radioactivity, The terms used in equations 1 to 4 are not nuclear instrumentation repeated under the calculation section of the Friedlander, G., Kennedy, J. W., and Miller, individual determinations unless required for J. M., 1964, Nuclear and radiochemis- clarification. try: New York, John Wiley and sons, 520 p. Glasstone, S., 1958, Sourcebook on atomic Glossary energy: second edition, D. Van Nostrand Confidence level. The stated probability, un- Co., 525 p. der the experimental conditions employed, Hogerton, J. F., 1963, The atomic energy that the value will be within the interval deskbook : New York, Reinhold Publish- indicated by the precision around the ing Go., 623 p. mean. Overman, R. T., and Clark, H. M., 1960, Radioisotope techniques : New York, Decay. The spontaneous radioactive trans- McGraw-Hill Book Co.,464 p. formation of one nuclide into a different nuclide or into a different energy state of the same nuclide. Every decay process has Compilations of data on a definite half-life. radioactivity and radiochemistry Dissolved. The sample is filtered through a Holden, N.E. and Walker,F. W., 1969, Chart 0.45 micrometer membrane filter and the of the nuclides :Schenectedy, N.Y.,Gen- filtrate analyzed. eral Electric Co. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 15

Lederer, C., Hollander, J. M.,and Perlman,I., Wahl, A. C.,and Bonner, N. A., 1958, Radio- 1968, Table of isotopes: 6th ed., New activity applied to chemistry : New York, John Wiley and Sons, 593 p. York,John Wiley and sons, 536 p. National Academy of Sciences, National Re- search Council, 1960-62, A series of pub- lications on radiochemistry of the ele- Radioactivity regulations and ments beginning with NAS-NS publica- safety tion no. 3001, The radiochemistry of Federal Water Pollution Control Administra- : Revisions of publications on tion (now Environmental Protection several elements have since been pub- Agency), 1968, Water quality criteria : lished. 214 p. U.S. Dept. of Health, Education and Welfare, Public Health Service, 1970, Radiological General Safety Committee of the Manufac- health handbook :445 p. turing Chemists ASSOC.,1954, Guide for safety in the laboratory :New York,Van Nostrand, 221 p. Radioactivity in the environment International Atomic Energy Agency, Vien- Adams,J.A.S., and Lowder, W.M., 1964, The na, 1960, Safe handling of radioisotopes, Natural radiation environment :William health physics addendum : Safety series March Rice University, University of no. 2, 120 p. Chicago Press, 1069 p. -- 1962, Safe handling of radioisotopes : Jacobs, D. G., 1968, Sources of tritium and Safety series, no. 1,159 p. its behavior upon release to the environ- Morgan, K. Z.,and Turner,J. E.,1971, Prin- ment : U.S. Atomic Energy Commission ciples of radiation protection :New York, TID-24635, 118 p. John Wiley and Sons, 598 p. Saenger,E. L., 1963,Medical aspects of radi- Radioisotope methods in ation accidents: AEC contract AT (30- hydrology 1)- 2106 with University of Cincinnati, U.S. Atomic Energy Comm., 328 p. International Atomic Energy Agency, Vien- na, 1967, Guidebook on isotopic tech- U.S.Dept. of Commerce, National Bureau of niques in hydrology :175 p. Standards, 1951, Control and removal of --- 1970, Isotope hydrology (Latest in a radioactive contamination in labora- series of symposia on isotopic tech- tories: NBS Handb. 48,24 p. niques in hydrology. References to earli- -- 1953, Maximum permissible amounts er symposia are given in the 1970 publi- of radioisotopes in the human body and cation), 623 p. maximum permissible concentrations in air and water : Report of subcommittee Radioc hem ica I una lytica I 2 of the National Committee on Radia- methods tion Protection,NBS Handb. 52,45 p. -- 1959, Maximum permissible body bur- Crouthamel, C. E.,1970, Applied gamma-ray dens and maximum permissible concen- spectrometry : second edition revised by trations of radionuclides in air and Adams,F. and Dams,R., Oxford, Perga- water for occupational exposure: Recom- mon Press, 747 p. mendations of the National Committee Larrukhina, A. K,.Malysheva, T. V., Pavlot- on Radiation Protection,NBS Handb. 69, skaya, F. I., 1967, Chemical analysis of 95 p. radioactive materials : Cleveland, CRC. U.S.Public Health Service, 1962, Drinking Press, a division of Chemical Rubber Co., water standards, 1962 : U.S. Public originally published in Moscow, 360 p. Health Service Pub. 956,61 p. 16 TECHNIQUES OF WATER-RESOURCESINVESTIGATIONS

U.S. Environmental Protection Agency, 1973, Jacobs, 0. G., 1968, Sources of tritium and its be- havim upon release to the environment: U.S. 1972 Water quality criteria, : EPAaR3.- Atomic Energy Comm. TID-24685,118 p. 73.033, U.S. Govt. Printing Office, Wash. Johnson, J. O., 1971, Determination of radium-228 D.C., 594 p. in natural water: U.S. Geol. Survey Water- 1975, Interim primary drinking water Supply Paper 1696-G, 26 p. Kaufmann, S., and Libby, W. F., 1954, The natural regulations: Title 40, Code of Federal distribution of tritium: Physical Review, V. 93, Regulations ; Part 141, Federal Register, p. 1337-1344. v. 40,no. 150,Aug. 14, 1975,5 p. Krause, D. P., 1959, Ra-228 (Mesothorium I) in U.S. Nuclear Regulatory Commission, 1976, Illinois well waters: Argonne Nat. Lab. Radiol. Rules and regulations: Title 10, Chapter Phy. Div. Semiannual Rept. ANL-6049, p. 51-52. 1, Code of Federal Regulations-Energy, Libby, W. F., 1955, Radiocarbon dating, second Part 20, Standards for protection edition: Univ. of Chicago Press, 124 p.

against radiation, Sept. 1975, updated Nydal, R., 1966, Variations in carbon-14 concentra- I through April 19,1976,18p. tion in the atmosphere during the last several years: Tellus 18, p. 271-275. Patterson, R. L., and Lockhart, L. B., 1964, Geo- graphical distribution of lead-SUO in ground level References air, in The Natural Radiation Environment: Chicago, Univ. of Chicago Press, p. 383-394. Barker, Johnson, Edwards, I(. and F. B., J. O., W., Pearson, F. J., 1965, Use of C-13/C-12ratios to Robinson, B. P., 1965, Determination of uranium correct radiocarbon ages of materials initially in natural waters: U.S. Ged. Survey Water- diluted by limestone: Internat. Conf. on Radio- Supply Paper 16964, 25 p. carbon and Tritium Dating, 6th, Pullman, Begemann, F., and Libby, W. F., 1957, Continental Washington, 7-11, June 1965, USAEC Conf. watesr balance, ground-water inventory and 650652, Proc., p. 357-366. storage times, surface oce'an mixing rates and Rama, M. K.,and Goldberg, E. D.,1961, Lead-210 world wide circulation patterns from cosmic in natural waters: Science, 134, p. 98-99. ray and bomb tritium, Geochim. Cosmochim. Rona, Elizabeth, Gilpatrick, L. O., and Jeffrey, L. M., Acta 12,p. 277-296. 1956, Uranium determination in sea water: Am. Brown, Eugene, Skougstad, M. W., and Fishman, Geophys. Union Trans., v. 37, p. 697-701. M. J., 1970, Methods for collection and analy- Scott, R. C., and Barker, F. B., 1959, Radium and sis of water samples for dissolved minerals and uranium in ground water of the United States: gases: U.S. Geol. Survey Techniques Water- Conf. on the peaceful uses of atomic energy, 2d, Resources Inv., book 5, chap. Al, p. 4-15. Geneva, Switzerland, 1958, Proc., v. 2, p. 154- Chalov, P. I., Tuzova, T. A., and Musin, Ya. A., 157. 1964, Isoptope ratio of U-234/U-238in natural Starik, I. E., 1959, Principles of radiochemistry: waters and utilization of the ratio in nuclear Publishing house of the Academy of Sciences, geochronology: Izv. Akad. Nauk. SSSR, Ser. USSR, Translation AEC-tr-6314, Office of Geofiz, no. 10 p. 1552-61. Tech. Services, Dept. of Commerce, Washing- Cherdyntsev, V. V., Orlov, D. P., Isabaev, E. A., ton, D.C. 20230. and Ivanov, V. I., 1961, in Suess, H. E., 1965, Secular variations of the cosrnic- natural conditions, Radiokhimijia, no. 10, p. ray produced carbon-14 in the atmomhere and 840-848. theh interpretations: Jour. Geophys. Res., 70 (23), 5937-5952. Cooper, J. A., Rancitelli, L. A., and Perkins, R. W., Thatcher,'i. L., 1969, Principles of the application 1970, An anticoincidenceshielded Ge(Li) of nuclear techniques to hydrologic investiga- gamma-ray spectrometer and its application to tions, The progress of hydrology: First In- radioanalytical chemistry problems :Jour. Radio- ternat. Seminar for Hydrology Professors, Univ. anal. Chem. 6, 147-163 p. of Illinois, Urbana, Proc., p. 149-193. Edwards, K. W.,1968, Isotopic analysis of uran- Thurberr, D. L., 1962, Anomalous U-234/U-238in ium in natural waters by alpha spectrometry: nature: Jour. Geophys. Research, v. 67, p. 4518- U.S. Geol. Survey Water-Supply Paper 1696-F, 4520. 26 p. Wood, W. W., 1976, Guidelines for collection and Holtzmann, R. B., 1964, Lead-210 and polonium-210 field analysis of ground-water samples for in potable waters in Illinois, in The Natural selected unstable constituents : U.S. Geol. Survey Radiation Environment: Chicago, Univ. of Techniques Water-Resources Inv., book 1, chap. Chicago Press, p. 227-238. D2, (In press). Carbon-14, dissolved, apparent age Liquid scintillation method, Denver Lab (R-1100-76)

Parameter and code: Carbon-1 4, dissolved, apparent age (years): none assigned

1. Application ~CZHZ+ CeH,. The method determines the apparent age The benzene is placed in a tared scintilla- of carbon-14 dissolved in the water sample. tion counting vial, weighed, and then diluted It is suitable for the analysis of any natural- slightly with a toluene solution containing a water sample from which 5 g of dissolved mixture of scintillators sensitive to low- carbon can be obtained. Only carbon in the energy betas. The activity of the sample is form of dissolved CO, and its hydrolysis measured in a liquid scintillation counter. products is determined. The apparent age, From the ratio of carbon-14 activity to the determined by comparison with a standard weight of carbon recovered,the apparent age of known age (see see. 5.10), applies only to of the carbon in the sample may be de- the carbon-14 in the sample; use of the de- termined. termination to date the water itself requires a number of additional measurements and 3. Interferences assumptions and a general knowledge of the geohydrologic system from which the sample The conversion from CO, t0 C,H6 is ap- was obtained. As noted below, a modification parently free of both chemical and radio- of the procedure is used for waters contain- metric interferences. During sample collec- ing sulfate in excess of approximately 200 tion and precipitation procedures interfer- ence may result from conrtact with the atmos- W/l. phere or from dissolved sulfate concentra- tions in excess of approximately 200 mg/l. 2. Summary of method The sample collection must minimize at- The method is based on that of Noakes mospheric contact with the water. Sample and others (1967). Dissolved carbon in the collection should be in a closed system from carbonate system is concentrated from a the sampling point to the collection contain- large volume of water by precipitation of ers. Collection containers (normally 15- to barium carbonate. The precipitate is treated 55-gal steel drums) should be filled by in- with acid to liberate carbon dioxide, which serting tubing to the container bottom, and is then allowed to react with metallic flow continued until three container volumes to produce lithium carbide: have passed through the container. The 2C0,+ lOLi + Li,C,+ 4Li20. sampling tube should be removed slowly and The carbide is then hydrolyzed to produce the container capped immediately after re- acetylene : moval of the delivery tube. LizC2+ 2H,O + 2LiOH + C,H,. Finally, the acetylene is passed over a 4. Apparatus -doped aluminum oxide catalyst to 4.1 Beakers, polyethylene, 400 ml. fom benzene : 4.2 Cylinder, graduated, 50 ml.

17 18 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

city with 1- and 2-inch bungs. 4.4 Liquid scintillation counter. 4.5 pH indicating paper, pH 2.0-5.0 range. 4.6 Precipitator for barium carbonate (fig.2). 4.7 Vacuum line for storage of acetylene and conversion to benzene (fig. 3). 4.8 Valves for transfer of water sample from drums to precipitator.

5. Reagents 5.1 Ascarite (trade name of Arthur H. Thomas Co. for hydroxide-impreg- nated asbestos). 5.2 Barium chloride- chloride solution: Dissolve 290 g BaC1,-2H,O and 10 g LaCI, in distilled water and dilute to 1 liter. 5.3 Benzene, spectroscopic grade, free of 11 carbon-14 activity. 5.4 Catalyst, vanadium-doped A1,0,, or Mobil Oil Co. Durabead. 5.5 Dry ice. 5.6 Isopropanol, technical grade. 5.7 Liquid nitrogen. 5.8 Lithium metal, shot, packed, and stored under . 5.9 Nitrogen gas. Must be free of CO, or scrubbed to remove CO,. 5.10 Oxalic acid. National Bureau of Standards contemporary standard. 5.11 Phosphoric acid, concentrated (85 percent). A - PLEXIGLASS3 VIEWING PORT 5.12 pentoxide, granular. B - CONNECTOR FOR HOSE 5.13 Scintillator solution: 10 g 2,5-di- FROM SAMPLE BARREL phenyloxazole (PPO), 0.25 g 1,4-bis-2-(4- C - STIRRING MOTOR methyl-5-phenyloxazolyl ) -benzene (dimethyl D - PRESSURE RELIEF VALVE POPOP). Dissolve in 250 ml analytical- E - IMMERSION HEATER grade toluene. F - MASON JAR G - GATE VALVE 5.14 Sodium hydroxide solution, 5 M. 5.15 Sulfuric acid, 0.0164 N. Figure 2.-Apparatus for collection of carbonates 5.16 Strontium chloride-lanthanum chlo- from a water sample. ride solution: Dissolve 230 g SrC12.2H,0and 10 g LaC1, in distilled water and dilute to 1 liter. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 19

H / TO WATER RESE R V 0I R COOLING WATER

A - CARBON DIOXIDE EVOLUTION FLASK L - CATALYST REACTION TUBE B, C, D, G, H, I, J, K, 0 - TRAPS M - COLD FINGER FOR BENZENE E-1 - E-4, N - GAS STORAGE CYLINDERS P, Q, R - COLD FINGERS F - REACTION CHAMBER G-1 - G-7 - VACUUM GAUGES

Figure 3.-Vacuurn line for preparation of acetylene and conversion to benzene.

6. Procedure the solution in the 400-ml beaker. Again, I test the pH. If the pH is now above 4.5, it 6.1 Determine the volume of water sample is satisfactory to use two 15-gal drums of required to colntain 5 g carbon as carbonate water sample. If the pH is still less than 4.5, or bicarbonate. Alkalinity contributed by repeat the addition of 50-ml sample aliquots silicate, borate, phosphate, and other basic to the 400-ml beaker until a pH of 4.5 is constituents is included in the following esti- obtained. Collect one 15-gal drum for each mation of carbonate alkalinity. Therefore, 50-ml portion of sample required for neu- the method underestimates the volume re- tralization of the acid to pH 4.5. Six 15-gal quired if noncarbonate alkalinity is also or two 55-gal drums is the maximum volume present. collected. 6.1.1 To 50-ml water sample in 400-ml 6.2 Collection of sample. beaker add 30 ml 0.0164 N sulfuric acid. Test 6.2.1 Fill each 15-gal drum by inserting the pH, using narrow-rangeindicating paper a hose to the bottom of the drum and fill or a pH meter. If the pH is above 4.5, suffi- until two or more drum volumes have over- cient carbon is contained in 15 gal (55 liters) flowed. Remove hose, insert plugs in bungs of the water sample. and tighten securely,taking care to minimize 6.1.2 If the pH is less than 4.5, add a trapped air. Ship to the laboratory taking second 50-ml portion of water sampIe to care to prevent freezing in cold weather. 20 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS 6.3 Collect dissolved carbonate species by 6.4.1.2 Evacuate the 4-liter flask care- 1 precipitation as barium carbonate using the fully until bubbles foam in the slurry. Con- precipitation apparatus shown in figure 2. tinue evacuation for 2 min. 6.3.1 Attach a 2-liter Mason jar to the 6.4.1.3 Carefully run the phosphoric bottom of the precipitation cone. Run a hose acid from the separatory funnel into the 4- from the 15-gal drum containing the sample liter flask. Carbon dioxide is released from to the precipitator. Seal the top plate and the slurry. Allow the pressure to build up sweep out the unit with nitrogen gas to re- to approximately 120 mm of move atmospheric carbon dioxide. Apply (gauge G-1) and then open the stopcock to pressurized nitrogen to the 15-gal drum to traps C and D so that pressure holds con- force the water sample over into the pre- stant. Trap B is cooled with isopropanol-dry cipitator. ice. Its function is to condense water. Traps 6.3.2 Precipitate barium sulfate and C and D are cooled with liquid nitrogen, and carbonate by heating the sample with the their function is to condense carbon dioxide. immersion heater while stirring and adding 6.4.1.4 When the evolution of carbon 2 liters of barium chloride solution. If sul- dioxide is completed as shown by decrease fate in the water sample exceeds approxi- of pressure on gauge G-1 to a constant mately 200 mg/l (previously determined in minimum, carbon dioxide can now be trans- the field or laboratory), strontium chloride ferred to the storage cylinders E-1, E-2, is used as precipitant. Add 5 M sodium E-3, and E-4. The quantity of carbon diox- hydroxide slowly until pH of 10.4 is reached ide is calculated from the known volume of to convert bicarbonate to carbonate. Stir for each cylinder (slightly more than 6 liters) 1 hr while holding the temperature at ap- and the pressure on gauge G-3. Remove the proximately 40°C. liquid-nitrogen Dewar from trap C. Replace 6.3.3 Open the bottom valve and permit the liquid-nitrogen Dewar at trap D with the barium sulfatecarbonate precipitate to an isopropanol-dry ice Dewar. Carbon di- flow into the Mason jar. Rotate stirring rod oxide flows into cylinder E-1 and when briefly by flipping control switch after 1 hr. atmospheric pressure is reached, cylinder Repeat cycle each one-half hour for next 2 E-2 is opened and filled, The remaining two hr. Major portion of precipitate should now cylinders are successively filled in the same be in the Mason jar. Close bottom valve, un- way. screw Mason jar, cap immediately, and seal 6.4.1.5 Calculate the number of moIes exposed cap edges with vinyl or rubber tape. of carbon dioxide collected using volume of 6.4 Synthesis of benzene. The vacuum line each storage cylinder, gauge pressure, and for evolution of carbon dioxide and its con- ambient temperature. Determine the grams version to benzene is shown schematically of lithium metal to be used in the following in figure 3. carbide conversion step by multiplying the 6.4.1 Evolution of carbon dioxide from number of moles by 60. barium carbonate. 6.4.1.6 Weigh out the required lith- 6.4.1.1 Decant most of the supernate ium shot, and place in the steel reaction from the barium sulfate-barium carbonate chamber F. Evacuate the line and the cham- precipitate. Slurry the remaining precipitate ber, start the flow of cooling water through and solution and transfer to a 4-liter heavy- the reaction chamber,and turn on the heater. walled Pyrex flask (A). Put a magnetic 6.4.2 Reduction of carbon dioxide to stirrer bar in the flask, cap with a two-hole lithium carbide carbide. stopper containing a separatory funnel, and 6.4.2.1 When lithium is a dull-red place on the vacuum line. The vacuum line heat (as observed through the viewing port) has been previously evacuated. Fill the admit carbon dioxide from the storage tanks separatory funnel with 125 ml of 85 percent into the reaction chamber F. Pressure drops phosphoric acid. sharply ax the reaction proceeds (G-4).Con- METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 21 tinue heating until pressure drops to min- gen Dewar with isopropanol-dry ice. Com- imum. Continue heating for 1 hr. Turn off plete reaction is indicated by conistant heat and allow to cool. minimum pressure on gauge G-6. Complete 6.4.2.2 The lithium carbide contain- reaction requires 2-3 hr. A second tube of ing carbon-14 is treated with water to form catalyst may be required to achieve complete acetylene. Unreacted lithium reacts with reaction. water to produce hydrogen. Carefully intro- 6.5 Preparation for counting. duce water from the distilled-water reser- 6.5.1 Transfer the catalyst tube contain- voir. The pressure rises because of produc- ing benzene to a vacuum-distillation appa- tion of acetylene and hydrogen. The ace- ratus. Immerse the receiving tube in iso- tylene condenses in traps I and K which are propand-dry ice, and after the system is cooled with liquid nitrogen. Acetylene is evacuated, heat the catalyst tube. When con- purified before condensation by passage densation of benzene in the receiving tube through traps G,H, and I. Traps G and H are is complete, remove and weigh to determine cooled with isopropanol-dry ice and trap I benzene recovery. contains ascarite and phosphorous pentoxide. 6.5.2 Transfer 3.0 ml benzene to a scin- One to 1.5 liter of water is required for com- tillation-counting container, add 1.0 ml scin- plete reaction. When the pressure nears at- tillator, and count under optimum conditions mospheric, open the chamber to the vacuum for carbon-14. Count each sample several pump to remove hydrogen as fast as it is times to accumulate at least 400 min of produced. Continue pumping after comple- counting time on each sample. Reject tion of the reaction (bubbling ceases) until early counting run results if instability is pressure falls to full vacuum. displayed. 6.4.3 Formation of benzene from acety- lene. 7. Calculations 6.4.3.1 The reaction tube L contains Calculate the apparent age of the sample approximately 150 g of catalyst. Remove the from the following equation : liquid nitrogen Dewars from traps J and K Ts3.32 T, (l~gAo-logA), and replace with isopropanol-dry ice coolant around K only. Place a liquid nitrogen where Dewar around the cold finger M. Acetylene T= apparent age of sample in years, now sublimes from J and K and condenses T,= half-life of carboa-14 in years in M. (5,568 yr), 6.4.3.2 Conversion of acetylene into A, = 0.950 times activity (in net counts benzene via the aluminum oxide catalyst par minute pel- gram of carbon) averages about 97 percent with a good batch of NBS oxalic acid contemporary of catalyst. Occasionally an inferior batch is standard. This value is an aver- encountered, and conversion is much lower. age of several measurements on There is no way to predict whether a new contemporary standards, and shipment of catalyst will give high or low A =!activity of sample (in net counb conversion. The percentage of conversion for per minute per gram of carbon). an individual run is determined by diverting the acetylene into storage cylinder N before 8. Report it is frozen into cold finger M. The pressure Report apparent age of dissolved arbon- reading shown on gauge G-6 and the known 14 in sample to nearest 50 yr for ages < 1,000 volume of N enables calculation of the moles and to nearelst 100 yr for ages > 1,000. of acetylene. 6.4.3.3 When acetylene is completely 9. Precision condensed in M transfer it to the catalytic Precision as calculated from the counting reaction tube by replacing the liquid nitro- variances for the sample, background, and 22 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS the standard comumtsgive an optimistic esti- References mation since possible errors involved in the Noakes, J. E., Kim, S., and Akers, L., 1967, Recent precipitation ot the carbon and its wnversion improvements in benzene chemistry forr zadio- to benzene are not included. According to carbon dating, Gemhim. Cosmochim. Acta 31, Shiver (1972), the precision of a carbon-14 (6) p. 1094-1096. Shiver, M., 1972, in Encyclopedia of gcnchemish-y date should be +- 100 yr in the 10,000-yr range and environmental sciences, edited by Fairbridge, and i. 800 yr in the 30,000-yr range. Van Nostrand Reinhdd Co., p. 131. Cesium-137 ond cesium-134, dissolved Inorganic ion-exchange method-gamma counting (R-1110-76)

Parameters and codes: Cesium-137, dissolved, (pCi/l): none assigned Cesium-1 34, dissolved (pCi/l): 2841 0

1. Application ma detector and an automatic sample chang- The combination of a reasonably specific er and single channel gamma spectrometer. ion-exchange separation of cesium isotopes The signal to noise ratio is optimized by with the energy discrimination available adjusting the spectrometer window to the through gamma spectrometry provides a cesium-137or ceisium-134 energy peaks. very specific method with potentially wide Blanks consisting of test tubes with ex- applicability. Boni (1966) applied the related changer are counted with the comparator KCFC technique to the determination of standards and samples for a minimum of cesium-137 in milk, urine, seamwater and three 50-min counts. The cesium concentra- freshwater. Petrow and Levine (1967) ap- tion is calculated on the basis of the net plied an ammonium hexacyanocobalt ferrate gamma counts observed in the standards and (NCFC) method to the determination of sampies. cesium-137 in precipitation. 3. Interferences 2. Summary of method Ruthenium, -niobium,, and The method is a development of Janzer, were reported to be sorbed by KCFC based on the work of Petrow and Levine from neutral aqueous solutions. So,rptionof (1967), who used NCFC for the wncentra- interfering radionuclides was reduced to less tion of cesium isotopes from water. The am- than 0.1 percent from a 10 N HCl and 0.5 N monium compound is superior tu the potassi- HF solution (Boni, 1966). Boni also noted um compound (KCFC) used by Prout, Rus- that interfering radionuclides could be re- sell, and Groh, (1965) because of elimination moved by passing the sample through 50-100 of background from potaslsium-40. mesh Bio-Rad Celex 100 in the form The gamma-countiag technique begins before collecting the cesium on the KCFC. with collection of radiocesium from relatively Ellenburg and McCown (1968) reported large volumes of water (up to 20 liters) by iodine-131 and -99 were sorbed passing the sample through a column of in- in the analysis of reactor-fuel solutions using organic ion-exchanger (NCFC). Several a slurry technique with 150 ml of sample and standards are prepared by passing a meas- 50 mg of KCFC sorber. NCFC probably ex- ured amount of standardized cesium-137 so- hibits similar sorption. The degree to which lution in water through columns prepared in extraneous radionuclides collected by NCFC the same manner as those used for the un- may interfere during counting is a function knowns. The columns are dried and then of the energy resolution of the detector. counted using a well-type sodium iodide gam- Iodine-131 might interfere using the NaI de- 24 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS tector specified, because the cesium-137 from 0.3 to 0.5 meq Cs/g KCFC using a cesi- photopeak at 662 keV might not be fully re- um nitrate feed solution varying from 0.001 solved from the iodine-131 photopeak at 636 to 0.008 M. They calculated a theoretical ex- keV. Using a Ge(Li) detector, there would change capacity of about 6 meq/g for KCFC. be no interference because resolution would 5.2 Cesium-137 and cesium-13.4 standard be complete. The combination of decay to solutions: Obtain from the National Bureau eliminate molybdenum-99 (T$=67 h) and of Standards or orther commercial suppliers. other short-lived nuclides and high-resolution Dilute to approximately 100 pCi/ml for use. gamma spectrometry would appear to permit specific determination of cesium isotopes. 6. Procedure 6.1 Using the slotted 16x150-mm test 4. Apparatus tubes, prepare exchange columns by placing 4.1 Gamma spectrometer, single-channel small plug of glass wool in bottom of tube, type. add 30-60 mesh NCFC to a depth of 1 cm, 4.2 Glass and plastic tubing. and position a porous polypropylene disk on 4.3 Glass wool. top of the NCFC to keep it in place. Do not 4.4 Porous polypropylene disks: Cut compress the NCFC granules, or excessive from 1.5 mm thick, 120 micrometer porous flow reduction will result from the slight hydrophilic polypropylene sheeting (Bel Art swelling which occurs when the sorber is F-1256 or equal). wetted. Prepare columns as uniformly as 4.5 Sample changer, automatic for 16 possible to obtain reproducible counting ge- x150-mm test tubes, coupled to a printout ometry in the well-type detector. system.. 6.2 Using-5-mm glass and plastic tubing, 4.6 Slotted 16x150-mm test tubes: A 2- prepare a series of siphons which will pro- to 3-mm vertical slot, 1-mm wide, is cut into vide a 1- to 1.6-meter head. Firmly insert the the bottom of the test tube. Fire-polishing lower end of the siphon into the numbered the tubes after cutting reduces breakage. test tube exchanger using a size 0, one-hole rubber stopper,and place the upper end of the 5. Reagents siphon in the water sample. Maximum sam- 5.1 Ammonium hexacyanocobalt ferrate, ple volume normally used is 20 liters. (NCFC) 30-60 mesh prepared after the 6.3 Apply suction to the slot of the test manner described by Petrow and Levine tube to start flow. Flow rates of 1-5 ml/min (1967). are normal. More rapid flow rates can be ob- Add 10 ml of 0.5 M sodium ferrocyanide tained by using comer NCFC or increasing solution dropwise (approximately 35 ml/ the siphon head, but corntact time would be min) to 240 ml of 0.3 M cobalt nitrate-1.0 reduced accordingly. Some color may be M ammonium nitrate solution while stirring leached out of the NCFC during the first few on a magnetic mixer. Centrifuge slurry in milliliters of flow, but this apparently has no 250-ml centrifuge tubes, decant, and discard effect on the cesium collection. the supernate. 6.4 When all the solution has passed Wash slurry with water several times to through the column, disconnect the siphon, remove excess unreacted salts. Dry slurry in wipe the tubes dry, and place in an 80°C oven tubes by heating overnight in 80°C oven. to dry. Crush dried salt and sieve to collect 30-60 6.5 Prepare several comparator standards mesh fraction. Retain fines for incorporation by passing a measured amount of cesium-137 with next batch of slurry prepared. standardized solution (20-100 pCi) through The measured cation exchange capacity additional exchanger tubes. for 30-60 mesh KCFC was reported by 6.6 Adjust the gamma spectrometer to Prout, Russell, and Groh (1965) to range obtain optimum counting in the cesium-137 METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 25 energy region (662 keV) or the cesium-134 is set equal to E, or Ea.Decay correction for energy region (605 keV,796 keV) . cesium-I37is usually not required. 6.7 Count reagent blanks, standards, and 1oooz samples for a minimum of three 50-min pCi/l of cesium-137= KVE,(e-V) * counting periods each. 1OOOT pCi/l of cesium-134= 7. Calculations KVEb(e-xbt) * 7.1 Efficiency factors for cesium-137 (Ea) 8. Report and cesium-134 (Eb)are calculated by the following equations. The efficiency factor in- Report concentrations to one significant cludes counting efficiency and chemical re- figure for concentrations between 0 and 10 covery. Use of standard equation 2 to correct pCi/l and to two significant figures for high- for cesium-137 decay is not often required. er concentrations. 9. Precision Precision of the method is estimated to be approximately 420 percent. where - References c, = average count rate of standard Boni, A. L., 1966, Rapid ion-exchange analysis of (cpm) corrected for background radiocesium in milk, urine, sea-water and en- and blank, vironmental samples: Anal. Chemistry, v. 38, d, = disintegration rate of standard no. 1, p. 89-92. Ellenburg, E. J., iand McCown, J. J., 1968, Rapid (dpm)t carrier-free method for the radicuchemkd deter- &= decay constant of cesium-137 mination of cesium-137: Anal. Letter5, v. 1, no. (0.02295 yr-l) , 11, p. 697-706. hb = decay constant of cesium-134 Petrow, H. G., and Levine, H., 1967, Ammonium (0.0280 months-') , and hexacyanocobalt ferrate as an (mproved inor- ganic exchange material for deknninlation of t, = elapsed time between certification of cesium-137: Anal. Chemistry, v. 39, no. 3, p. the standard and the count, in 360-362. same units as the respective A. Prout, W. E., Russell, E. R., and Groh, H. J., 1965, Ion exchange absorption of crssium by potas- 7.2 Calculation of cesium-137 and cesium- sium hexacyanacabalt (11) ferrate (11) : JOUT. 134 concentrations :Use equation 1 where Ef of Inorg. Chem., v. 27, p. 473-479.

Radiocesium, dissolved, as cesium-137 Inorganic ion-exchange method-beta counting (R-1 1 11-76)

Parameter and code: Radiocesium, dissolved, as cesium-1 37 (pCi/l): none assigned

1. Application The method determines total dissolved Application is possible when identification radiocesium concentration because individual of individual cesium isotopes is not required isotopes are not identified by this beta count- and when interfering beta-emitting isotopes ing. are in low concentration. Concentration limits for interfering isotopes have not been 3. Interferences fully evaluated but would appear to be lower No interferences have been found. than for the gamma-spectrometry technique for cesium-137 and cesium-134, dissolved 4. Apparatus (inorganic ion-exchange method-gamma 4.1 Low-background counter, an anti- counting, R-1110-76). Until interferences coincidence-type counter with 2-in.thin win- are quantitatively evaluated, examination of dow flowing gas propcdoaal detector pref- each sample for possible interference is ad- erably capable of measuring both alpha and vised. Applications to samples where radio- beta activity simultaneously. (The Beckman cesium is the principal source of radioactiv- Wide Beta or Low Beta counters or equiva- ity would appear to be safe. lent, with pulse height discriminating cir- cuitry, are satisfactory.) 2. Summary of method 4.2 Filter, nitrocellulose membrane filters, In the beta-counting technique (Janzer, 0.45-micrometer pore size, 47-mm diameter. 1973) the same ion-exchanger (NCFC) is A suitable filter holder assembly to facilitate used to collect the cesium-137 and cesium- vacuum filtration using the 47-mmfilters is 134 isotopes as in the gamma technique, required. The perforated backup plate under but a simpler batch operation is made pos- the filter disk must be plastic rather than sible by the smaller volume of water used. glass because of the hydrofluoric acid solu- One hundred milligrams of the ion-exchanger tion used. Millipore Sterifil XX-047-10 or are stirred with the buffered water sample equivalent is satisfactory. and then separated by filtration through a 4.3 Teflon, polyethylene, polypropylene, or paper or membrane filter. This forms a uni- other acid resistant beakers, 600-mlvolume. form low-density disk deposit that is opti- mum for beta counting. Standards are pre- 5. Reagents pared using the same technique. The sample 5.1 Ammonium hexacyanocobdt ferrate and standard disks are counted in a low- (NCFC): Prepare as described by Petrow background beta counter with anticoinci- and Levine (1967), and sieve to collect the dence shielding. 30-60 mesh fraction. 28 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

5.2 Cesiwrn-137 standard solution. Obtaim If different sample sizes are used, corres- \ from NBS or use commercial standards, ap- ponding standards should.be prepared. proximately 100 pCi/ml. 6.7 Prepare blanks by running the deter- mination on 500 ml of distilled water. 6. Procedure 6.8 Coumt the blanks and standards in the 6.1 Place a 500-ml water sample in a plas- same way as the samples. tic beaker. 7. Calculations 6.2 Acidify with concentrated hydro- 7.1 The cesium-137 efficiency factor (E) chloric acid to 1 N and with hydrofluoric includes ohemical recovery and counting- acid to 0.5 N. For a 500-ml sample, this re- efficiency corrections and is determined by quires 45 ml of concentrated hydrochloric equation 2 simplified by treating standards and 10 ml of concentrated hydrofluoric acid. and samples the same so that no factor f is 6.3 Add 100 mg (21 mg) of 30-60 mesh calculated. The following equation is then NCFC, and stir for 10 min. After allowing used : the NCFC to settle, decant the supernatant solution and filter it through a 0.45-micro~m- eter membrane filter (47-mm diameter) in a vacuum-filtration assembly. With the last 7.2 Calculation of cesium-137 concentra- portion of liquid remaining in the beaker, tion: Use general equation 1 simplified by transfer the NCFC to the filter taking care eliminating the f term. Decay correction is to obtain even distribution. Return the fil- seldom reauired.- trate to the plastic beaker. Filter through 100OE pCi/l of &urn-137 = the original filter disk a second time, and KVE (,-At) ' quantitatively transfer any remaining NCFC 8. Report to the filter disk with small amounts of dis- ) tilled water. Rinse down the funnel sides, Report cesium-137 activity to one signifi- and carefully wash the filter and retained cant figure below 10 pCi/l and to two signifi- NCFC several times so that the filter pad is cant figures above 10 pCi/l. The minimum not disturbed and a uniform deposit is main- concentration reported is that which repre- tained. sents two standard deviations above back- 6.4 Mount the filter disk in a ring holder, ground. This is approximately 1 pCi/l with a and dry under an infrared lamp. Cover the 500-ml sample and a Beckman Low Beta dry material with a small sheet of plastic counter or equivalent. wrap (kitchen-type) . 9. Precision 6.5 Count in the anticoincidence beta counter. Three 50-min counts for each sample On the basis of limited data the analysis are usually adequate. appears to be reproducible to 410 percent at concentrations above 10 pCi/l with inferior 6.6 Prepare standards in triplicate. Add reproducibility at lower concentrations. 5.0 ml of 100 pCi/ml staindard cesium-137 solution to 500 ml of distilled water, and car- Reference ry out &he analytical procedure exactly as Janzer, V. J., 1973, A rapid mathod for the de- with a sample. The standards prepared in termination of radioactive cesium isotopes in this way are also mvered with plastic film water: U.S. Geol. Survey Jour. Research, v. 1, and may be rehined for semipermanent use. p. 113-115. Gross alpha and beta radioactivity, dissolved and suspended Residue method (R-1120-76)

Parameters and codes: Gross alpha, dissolved as U natural (pg/I): 80030 Gross alpha, suspended, as U natural (pg/l): 80040 Gross alpha, suspended, specific activity as U natural (pg/g): 01518 Gross beta, dissolved, as cesium-1 37 (pCi/J): 0351 5 Gross beta, dissolved, as stronti um-90/-90 (pCi/l): 80050 Gross beta, suspended, as cesium-1 37 (pCi/l): 0351 6 Gross beta, suspended, as strontium-90/yttrium-90 (pCi/l): 80060 Gross beta, suspended, specific activity as cesium-137 (pCi/g): 03518

1. Application and strontium-90/yttrium-90and cesium-137 The method is applicable to any natural- calibration standards, and results are re- water sample. Because of restrictions on the ported relative to these reference isotopes. weight of residue which can be accommo- Thus, the mealsured sample activity is re- dated, the sensitivity falls off with increasing ported in terms of the mount of natural concentrations of dissolved solids. uranium and equilibrium strontium-90/yttri- I um-90 and cesium-137 activity which would 2. Summary of method give the same alpha and beta count rates respectively for the same weight of residue. The method is an extension of the proce- The gross activities are reported in terms of dure published by Barker and Robinson equivalent quantities of reference standards (1963) for gross beta radioactivity. of the true alpha and beka activities of the A representative aliquot, but not more sample. than 1 liter of the sample including sus- pended solids, is filtered through a tared The accuracy of these approximations de- 0.45-micrometer membrane filter. The filter pends on a number of variables related to the and retained solids are dried at room tem- energy distributions of the alpha and beta perature and then at 105"C, cooled, and re particles and the similarity of the residues weighed to determine the weight of nonfiltra- used in preparation of the calibration curves ble residue per liter. to the actual sample residue. The method A filtered volume of the sample containing must be regarded as a rapid, semiquantita- no more than 150 mg of dissolved solids is tive measure of gross sample activity. evaporated to dryness in a Teflon evaporat- ing dish. The residue is transferred to a 3. Interferences tared, 2-in. concentric-ring, stainless-shl planchet, dried in a desiccator, weighed, and Within its intended purpose, the method is counted on a low-background alpha-beta free of interferences, although the accuracy counter. The observed radioactivity is com- varies considerably with the nature of the pared with the activity of natural uranium alpha and beta emitters, chemical composi- 30 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS tion of the sample, and uniformity of plan- 5.6 Uranium standard solution, 1.00 ml chet preparation. =lo0 pg U: Dissolve 0.1773 g of UO, (C,H,O,)2. 2H20 in approximately 500 ml 4. Apparatus distilled water. Add 15 ml concentrated 4.1 Evaporating dishes, Teflon, 100 ml. HNO, and dilute to 1,000 ml in a volumetric 4.2 Hotplate or steam table. flask. Store in a Teflon bottle. 4.3 Infrared drying lamps. 6. Procedure 4.4 Low-background counting equipment. Proportional counters capable of measuring 6.1 Preparation of beta calibra~oncurve. both alpha and beta activity are desirable 6.1.1 Add the following amounts of cali- (for example, Beckman Instrument Co. bration solutions A and B to 100-ml Teflon WidecBeta o'rLow-Beta 11, or equivalent). evaporating dishes : 4.5 Membrane filters, 47-mm diameter, Dish No. 0.45-micrometer pore size, cellulose nitrate (mff) or mtate type. 1 ______25 10 4.6 Planchets, stainless steel, 2-in. diame- 2------5 20 3 ______80 30 ter, concentric-ring type. 4------10 40 4.7 Specific conductance meter. 5 ______140 50 4.8 Vacuum desiccator. 6------20 60 4.9 Vacuum-filtration apparatus, for 47- 7 _-______- 200 70 mm membrane filter% 8------25 80 9 ______250 90 lo------_ 30 100 5. Reagents 11 -______300 110 12------35 121) 5.1 Calibration solution A :Dissolve 0.284 13 ______350 130 g 0.070 g NaCl, 0.026 g 14------40 140, MgS04-7H,0, I CaS04.2H,0, 0.109 g NaHCO,, and 0.245 15 ______400 150 g CaCO, in distilled water, bubbling CO, gas To each dish and to four additional 100-ml through the solution if necessary to obtain Teflon evaporating dishes (Nos. 16, 17, 18, clear solution. Dilute to1 2.00 liters. and 19) add 1.00 ml of strontium-90/yttri- 5.2 Calibration solution B :Dissolve 1.350 um-90 standard solukion. To dishes 18 and 19 g MgS04-7H,0,3.510 g NaC1, 1.550 g also add 5 drops concentrated NH40H. CaS04-2Hz0,0.508 g MgCl,.GH,O,and 0.300 6.1.2 Evaporate all the solutions to dry- g CaCO, in distilled water, using CO, bub- ness on a low-temperature hotplate. When bling as necessary to dissolve. Dilute to 2.00 dishes 18 and 19 are dry, raise heat to ap- liters. proximately 350°C to volatilize NH,CI. 6.1.3 Pdorm steps 6.6 through 6.11 of Note :Compocsition of the calibration solu- procedure which follows. tions should approximate that of the samples the to be analyzed. Solutions A and B were se- 6.1.4 Plot the be& efficiency (net counts per minute per picocurie) against the residue lected to approximate the composition aver- age of 12 major rivers in the United States. weight to obtain the beta calibration curve. 6.1.5 Prepare a beta calibration curve 5.3 Cesium-137 standard solution, ap- for cesium-137 following steps 6.1.1 through proximately 500 pCi/ml and acidified to 6.1.4. Use 1.00 ml olf cesium-137standard. approximately 1 N with hydrochloric acid. 6.2 Preparaioa of alpha calibtration curve : 5.4 Hydrofluoric acid, 49 percent. The alpha calibration curve is obtained in 5.5 Strontium-PO/yttrium-90standard so- exactly the same manner as the beta curve lution, approximately 500 pCi/ml combined (step 6.1) except that 100 pg of uranium (1 activity and acidified to approximately 1 N ml of uranium standard solution) is substi- with hydrochIoric acid. tuted for the strontium-90/yttrium-90 beta METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 31 standard. It is important that the uranium solids for the equivalent strontium-90/fltri- standard be in secular equilibrium with re- um-90 or cesium-137may also be calculated. spect to uranium-234. The uranium isotopes 6.6 Gross radioactivity of the dissolved ratio may be determined by the method de- solids is determined as follows :Using a tared scribed elsewhere in this manual. Teflon dish, evaporate the required volume of 6.3 Sample analysis :Measure the specific water to drynem on a hotplate or steam bath. conductance of each water sample. Multiply Remove Teflon dishes from hotplate as soon the specific conductance in pmhoa/cm at 25°C as they reach dryness to prevent warping of by 0.65 to obtain an approximate value for dishes due to excess heat. the dissolved-solidsconcentration of the sam- 6.7 Determine approximate weight of ple in mg/l. Determine the volume of sample residue by weighing dish plus residue and which will contain approximately 100 mg of subtracting tare weight. If the weight of dissolved solids : residue falls outside the 50- to 130-mgrange, 100 mg start a new evaporation using a larger or Sample volume, V (1) = mg/l smaller sample volume as required. The actual sample residue should weigh 6.8 Quantitatively transfer the residues between 50 and 130 mg. in the evaporating dishes to tared planchets 6.4 Gross radioactivity of the suspended using rubber policemen and a minimum solids is determined as follows: Vigorously amount of distilled water to effect the trans- agitate the sample bottle containing the total fers. Confine the solution and residue to the amount of sample collected and quickly pour three inner concentric rings of the planchet off 1 liter of sample with suspended solids during the transfer. into a graduated cylinder. Allow sediment to 6.9 Dry the planchets under infrared heat settle and then, ulsing suction, filter through lamps. Police down the evaporating dishes a tared membrane filter. When only 100 ml of with small additional amounts of distilled sample remains to be filtered, swirl to sus- water and transfer to planchets. pend solids and add suspension to filter fun- 6.10 Repeat step 6.6. If a residue remains nel. Any remaining traces of sediment may in the evaporating dish after three washes be transferred using small amounts of dis- with distilled water, add a small amount of tilled water. hydrofluoric acid, and use for a final policing. 6.5 Carefully rinse solids with a small 6.11 Disperse the final liquid slurry in the amount of distilled water, and maintain as planchets as uniformly as possible, using a uniform a deposit as possible. Remove filter, stirring rod to disrupt large aggregates. allow to air-dry, and weigh. If the weight of Again evaporate to dryness. solids is in excess of 150 mg, refilter a fresh 6.12 Place the planchets in a vacuum aliquot of appropriate volume to obtain less desiccator and evacuate to complete the dry- than 150 mg,weigh, dry at 105"C,and count ing; weigh the planchets and determine the for alpha and bets radioactivity. weight of residue. Although not totally comparable,the same 6.13 Count the planchets in a low-back- absorption or counting efficiency factors for ground alpha-beta counter. Obtain three 50- the respective counting instruments are used min counts on each sample. as for dissolved solids, and the calculations are made in the same way. The weight of sus- 7. Calculations pended solids per liter of sample is also re- ported (USGS parameter code 00530) and Determine the equivalent concentration of may be used to calculate the radioactivity as- alpha and beta emitters in each planchet sociated with the toital weight of suspended from the net measured activity and the ap- solids per liter of original sample. Specific propriate factor from a table of calibration activity of the solids in pCi/g of suspended curves. 32 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

Scpmff 1000 8. Report Gross a=- X-=pg/l as U natural, Fa Va Report values of less than 1 pCi/l to one FB * 1000 significant figure. Report higher values to Gross p = Scpmp x V, two significant figures. =pCi/l as SrS0/Yg0or as CsI3: where 9. Precision Scpm= alpha count rate of sample in Results for a particular sample are usually reproducible to about a20 percent at the 95- counts per minute, - Scpmp=beta count rate of sample in percent confidence level. counts per minute, Fa= alpha factor in cpm/pg of natu- ral U, Reference Fp=beta factor in picocuries per Barker, F. B., and Robinson, B. P., 1963, Debmina- count per minute, and tion of beta activity in water: U.S. Geol. Survey V,= aliquot used in milliliters Water-Supply Paper 16964, 32 p. Lead-210, dissolved Chemical separation and precipitation method (R-1130-76)

Parameter and code: Lead-210, dissolved (pCi/l): none assigned

1. Application and other natural radionuclides do not inter- The method may be used with all natural fere. Lead-212 does not interfere because of waters where detection of lead-210 with sen- the short half-life (10.6 hr). Stable lead in sitivity to 2 pCi/l is acceptable. Where im- the water sample should cause no interfer- proved sensitivity is required, a larger sam- ence by increased beta absorption since lead ple than the normal 500 ml is concentrated concentrations seldom exceed a few tenths of by evaporation before beginning the pro- a milligram per liter. A problem might be en- cedure. Rama and Goldberg (1961) reported countered with industrial wastes where lead they were able to,achieve sensitivity to 0.02 concentrations could be much higher. pCi/l by use of a 20-liter sample and direct The application of the method to liquid precipitation of lead chromate from this waste should be checked by “spiking” sam- large volume. ples with known lead-210 and determining Lead recovery may be low in waters con- percentage of recovery. taining high concentrations of organic ma- terial that can possibly complex lead. 4. Apparatus 4.1 Centrifuge, capable of handling 50-ml 2. Summary of method tubes. The analytical method is designed to iso- 4.2 Centrifuge tubes, polypropylene, ap- late lead-210 in a relatively pure lead chro- proximately 50-ml capacity. mate precipitate. This is allowed to age to 4.3 disks, 49-mm diameter, mini- produce bismuth-210,a beta emitter with 5-d mum thickness 0.1 mm. half-life. Lead-210 is not counted directly bc+ 4.4 Ion-exchange columns, 1-cm inside di- cause of its very soft beta radiations of 15 ameter, 10-cm long with 50-ml reservoir at and 61 keV,which are greatly attenuated by top4 ab,sorption.The bismuth isotope decays by 4.5 Low-background counter, an anticoin- beta emission with a beta maximum energy cidence-type counter with 2-in. thin window of 1.16 MeV, and these beta particles are flowing gas proportional detector preferably easily counted. capable of measuring both alpha and beta activity simultaneously. 3. Interferences 4.6 Magnetic stirrer-hotplate. 4.7 Membrane filters, 47-mm diameter, Henry and Loveridge (1961) have shown 0.45-micrometer pore size. Must be inert to that essentially complete Separation from warm, dilute chromic acid. radioisotopes of , cesium, cobalt, 4.8 Planchets, stainless steel, 5-cm iodine, phosphorous, ruthenium, strontium, diameter. , zinc, and zirconium is achieved. 4.9 Tape, celluloee, 54m wide roll, adhe- Radium, thorium, uranium, potassium-40, sive on both sides.

33 34 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

4.10 Tape, “Magic Tape” (Minnesota 6.1.2 Plug the end of the column with a I Mining and Manufacturing Co.), 8 mg/cm2 small piece of glaes wool, and wash with 8 N density, 5-cm width. HC1. 4.11 Vacuumfiltrationapparatus, for 47- 6.1.3 Transfer the resin slurry com- mm membrane filters. pletely to the column, allow to drain and 5. Reagents settle. 5.1 Acetic acid, 25 percent by volume :Di- 6.1.4 Place a small glass-wool plug at lute 1 volume glacial acetic acid with 3 vol- the top of the resin column. Avoid packing umes distilled water. the resin. Wash with 10 column-volumes each 5.2 Ammonium persulfate. of 8 N HCl, distilled water, and finally 2 N 5.3 Hydrochloric acid, 8N and 2 N. HC1. The column is now ready for use. The 5.4 Ion-exchange resin, Bio-Rad Agl-X8, resin is dilscarded after one use. or equivalent, 50-100 mesh. 6.2 Preparation of lead-210 standard 5.5 Lead carrier solution, 1 ml=4.00 mg planchets : Three or more “permanent” Pbf2: Dissolve 6.394 g anhydrous Pb (NO,) standards are prepared by precipitating a in 100 ml of 1 N HNO,and dilute to 1,000 ml known amount of lead-210 as lead chromate. with distilled water. 6.2.1 Add 5 ml of lead carrier solution 5.6 Lead-210 standard solution. Suitable and an accurately known amount of lead-210 standards are available from Amersham/ (approximately 200 pCi is convenient) to 50 Searle Corp., Arlington Heights, Ill. Solu- ml of 0.001 M nitric acid. Heat to 95”C,slow- tions must be kept strongly acidic (about 1 N ly add 5 ml of 4 percent potassium dichro- in HNO,) when dilutions are made. mate solution, stir for several minutes, and 5.7 Nitric acid, 0.01 M: Dilute 0.1 ml of digest on a hotplate for 20 min with occasion- concentrated HNO,to 120 ml. al agitation. 5.8 Nitric acid, 0.001 M: Dilute 20 ml of 6.2.2 Cool to room temperature, filter 0.01 M HNO,to 200 ml. through a tared 0.45-micrometer membrane 1 5.9 Potassium dichromate solution, 4 per- filter, and wash the precipitate with a small cent :Dissolve 40 g reagent-grade K2Cr20,in amount of distilled water containing a few distilled water and dilute to 1 liter. drops of aermol OT solution. 5.10 Sodium acetate solution, 10 percent: 6.2.3 Allow the membrane and precipi- Dissolve 100 g anhydrous NaC,H,O, in dis- tate to air-dry,then weigh and calculate the tilled water and dilute to 1,000 ml. chemical yields. Mount the standards as de- 5.11 Sodium carbonate solution, 1.25 M: scribed in step 6.16. Dissolve 265 g of anhydrous Na,CO, in dis- 6.2.4 Allow the standards to age until tilled water and dilute to 2,000 ml. the count rate becomes constant. Thie will re- 5.12 Sodium sulfate, anhydrous. quire 40 d (8 half-lives of bismuth-210) or 5.13 Strontium carrier solution, 1 ml= 60 less, because 75 to 100 percent of the bis- mg Sr+,: Dissolve 145 g Sr(NO,), in dis- muth-210 coprecipitates with the lead tilled water; add 1 ml concentrated HNO, chromate, and dilute to 1,000 ml. 6.3 Analysis of the water samples: Meas- ure a 500-ml water sample into an 800-ml 5.14 Surfactant, Dade Aerosol OT, Scien- tific Products Co. beaker, add a magnetic stirring bar and a few drops of methyl orange indicator, and 6. Procedure place on a magnetic stirrer-hotplate. 6.4. While stirring, add 0.01 M nitric acid 6.1 Preparation of ion-exchange columns. dropwise until the indicator turns red. If 6.1.1 Slurry 4.0 g of ion-exchange resin the sample is initially acid to methyl orange, with 10 ml of 8 N HC1 in a 50 ml beaker. bring to color change with ammonia, then Cover and place in an ultrasonic cleaner for back to acid side with nitric acid. Add 5 ml 1 min and allow to stand for at least 20 min. of lead carrier solution, 4.5 g of anhydrous METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 35

sodium sulfate, and 1 g of ammonium persul- lecting the eluate in a 150-ml beaker. Record fate. Cover with a watch glass, and heat and the date and time of elution. This fraction stir at the incipient boiling point for 1 hr. contains lead-210 free of daughter activity. 6.5 Add 5 ml of strontium carrier solution 6.13 Add 3 ml of 25 percent (v/v) acetic to the sample and continue heating while acid and 5 ml of 10 percent sodium acetate .e0 stirring at the incipient boiling point for 20 the eluate, and dilute to approximately 75 ml. min ; retrieve the magnet, remove the beaker Heat to incipient boiling and add slowly 5 ml from the hotplate, and allow the precipitate of 4 percent potassium dichromate solution. to settle and solution to cool slowly to room Mix thoroughly by swirling the beaker and temperature. hold near the boiling point for 20 min with 6.6 Decant or syphon off as much of the occasional swirling. supernatant liquid as possible without dis- 6.14 Remove the beaker frotm the hotplate turbing the precipitate. Wash the residue and filter, while warm, through a tared 0.45- into a 50-mlround-bottom Teflon or polypro- micrometer, 47-mm-diameter membrane fil- pylene centrifuge tube using a mall volume +m.Rinse with distilled water, containing a of distilled water. Centrifuge, and discard the few drops of Aerolsol OT solution, to prevent supernate. the precipitate from clinging to the filter funnel. 6.7 Place a %-in. Teflon-coated magnet in 6.15 Remove the filter membrane from the centrifuge tube, add 40 ml of hot 1.25 M filtration apparatus, place on a clean sur- sodium carbonate solution, and place the tube the face, and allow to air-dry.Weigh the filter in a boiling-water bath on a magnetic stirrer- to determine chemical recovery. hotplate. Stir vigorously for 20 min. 6.16 Place the filter in the center of a 49- 6.8 Remove the tube from the bath, and mm copper disk, the surface of which wash down any precipitate on the upper part has been covered with a strip of double-faced j of the tube using a jet of 1.25 M Na2C0,. cellulose tape. Cover the filter disk with a Centrifuge and decant o’rsyphon off most of strip of 5-cm “Magic Tape,” trim to the size the supernate. Syphoning is preferable be- of the copper disk, and place in a 5-cm cause of the difficulty of decanting without planchet. disturbing the precipitate with the magnetic stirrer. 6.17 After an aging period of 7 d (or longer), count the samples in a low-back- 6.9 Repeat steps 6.7 and 6.8 two more ground beta counter with a 5-em detector for times. Remove the stirring bar before the 100 min to attain sensitivity to 2 pCi/l. Ex- final centrifugation. Discard the final super- tended counting improves the detection. For natant liquid. example, detection to 0.7 pCi/l i’s possible 6.10 Dissolve the lead-strontium carbon- with a 1,000-mincount. ate precipitate-in the centrifuge tube by cau- tiously adding 15.0 ml of 2 N HC1 followed 7. Calculations by 1.0 ml of 8 N HC1. Mix thoroughly and pour the solution onto the ion-exchange col- 7.1 Lead-210 efficiency factor (E) and umn. Be sure the flow rate does not exceed 2 chemical recovery factor (fn) :Although bis- ml/min. muth-210, daughter of lead-210,is actually counted, an in-growth factor is not used in 6.11 Rinse the centrifuge tube with two 2- the determination 02 E. The standard is al- ml prtions of 2 N HC1, adding these rinse lowed to reach equilibrium (40-d standing) solutions to the column. before counting, and the disintegration rate 6.12 Wash the column with a volume of 2 at this time is controlled by the lead-210 con- N HC1 equal tu two column-volumesplus the centration. Determine fn from the weight of holdup volume in the column below the resin lead chromate. E is determined using equa- bed. Elute with 30 ml of distilled water, col- tion 2. 36 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

1 equation 3 with the terms for decay of I daughter before counting and decay during 7.2 Calculation of lead-210 concentration : counting eliminated. Calculate lead-210 con- An in-growth factor is used in calculation of centrations using the equation below, and concentration in individual samples because correct for decay if excessive time elapsed the aging time before counting (approxi- between the sample collection and analysis. mately 7 d) is insufficient to attain equilibri- lO0OC pCi/l of lead-210=- f um. The in-growth factor for each sample is VEf (e-At) (1 -exA) where obtained from a curve of the growth of ac- - tivity with time (fig. 4) or, more accurately, e= net count rate of sample after cor- by calculation. Chemical recovery factor is rection for background and blank determined from the ratio of actual weight of A= decay constant of lead-210 (0.0315 lead chromate recovered to the theoretical Yr-l) 9 weight. For 20-mglead carrier, the theoreti- t = elapsed time between collection of cal weight is 31.20 mg of lead chromate. Use sample and analysis,

I- >- I- $ .8 4

5 .5

LL -3 .2 .1 2 13 14 123 45 6 78 91011 344 g-4E TIME IN DAYS Figure 4.-Growth of bismuth-210from pure lead-210source. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 37

A, = decay constant of bismuth-210 averaging 98 percent with a mean deviation (0.1384 d-,), of -+2 percent at concentrations above 15 t, = elapsed time between elution of the pCi/l. At lower concentrations the precision ion-exchange column (step 6.12) is +-2 pCi/l or a10 percent, whichever is and counting, and larger, V, E,and f are as defined in equation 1. 8. Report Repo'rt Concentrations of less than 10 References pCi/l to the nearest pCi/l ; higher concentra- tions to two significant figures. Henry, W. M., and Loveridge, B. A., 1961, Deter- mination of lead-210 in Harwell effluent: Atamic 9. Precision Energy Rwearch Estab. Rept. R-3795, 20 p. Tests with eight water samples to which Rama, M. K., and Goldberg, E. D., 1961, Lead-210 in known lead-210 was added gave recoveries natural waters: Science, 134 p. 98-99.

1 i Radium, dissolved, as radium-226 Precipitation method (R-1140-76)

Parameter and code: Radium, dissolved, as radium-226 (pCi/l): 09510

1. Application Radio.active impurities in the reagents may The method is satisfactory for applica- be significant. Reagents are selected, and tions that do not require high precision or blanks are run with each set of samples. radium isotope identification such as routine Since alpha radiation is strongly absorbed monitoring for compliance with PCRE by the precipitate, it is essential to keep the standards. The application is &raightfor- weight of precipitate constant in the samples, ward with waters of average composition, standards, and blanks. High salinity but the possibility of increased alpha self- (brines), colloidal matter, and high concen- absorption must be considered with waters trations of strontium or barium may cause of high dissolved-solids or alkaline-earth con- problems by adding to the mass per unit area tent. The internal standard technique is used of the precipitate. Correction for increased when necemary. absorption is made by use of internal stand- ard techniques. 2. Summary of method 3. Interferences Radium isotopes are concentrated from a water sample by coprecipitation as the sul- Alpha-emitting nuclides of thorium and fate with barium sulfate. The collection of polonium may be carried down with the pre- radium is quantitative even though the solu- cipitate, but their concentrations are not bility product of radium sulfate is not ex- sufficiently high in most natural water to con- ceeded. Although the coprecipitation collects stitute a problem. all radium isotopes, radium-226is usually the predominant isotope. Results are consequent- 4. Apparatus ly reported as concentration of radium-226. 4.1 Beaker, 1,500 ml. , lead, and polonium car- 4.2 Filter flask,2,000 ml. ried down partially or completely by the bari- 4.3 Membrane filter, 47-mm diameter, um sulfate may also contribute to the alpha 0.45-micrometer porosity. count. The analytical procedure is taken from Barker and Johnson (1964). 4.4 Filtration assembly, folr membrane fil- The precipitate is collected by filtration ter including funnel, sintered glass-filter sup- through a plastic-membrane filter, and is port and clamp. counted in a low-background thin window 4.5 Ring-and-disc sample mounts, 47-mm alpha counter after a suitable delay to allow diameter, in-growth of alpha-emitting daughters. This 4.6 Low-background counter, an anticoin- increases the sensitivity by a factor of 4 (at cidence-type counter with 2-in. thin window equilibrium). The count rate of the sample is flowing gas proportional detector preferably compared against that of a radium-226 capable of measuring both alpha and beta standard carried through the procedure. activity simultaneously.

39 40 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

5. Reagents and dilute to volume with distilled water. _I 5.1 Ammonium sulfate solution (40 per- Add a small, clean disk or foil of 22-carat cent, by weight) : Dissolve 1,600 g of am- (surface area E 4 em2)to the solution monium sulfate in a minimum volume of hot to remove polonium-210. This disk should distilled water. Cool to room temperature and remain in the standard solution continu- filter. Dilute the solution to approximately 4 ously. liters in a 4-liter bottle. 5.7 Sulfuric acid wash solution: Dilute 5.2 Barium carrier solution (1.40 mg 5 ml of concentrateld HSO, to 1,000 ml with Batz per ml) : Dissolve 1.246 g of barium distilled water and add 3 to 4 drops of Triton chloride dihydrate in distilled water, add 5 X-100 surfactant (Rohm and Haas Co.) . ml of concentrated HCl,and dilute to 500 ml. 6. Procedure 5.3 Hydrochloric acid, concentrated. 5.4 Methyl orange indicator solution. 6.1 Measure 1,000 ml of the sample, pre- 5.5 Radium standard solution I, 50.0 viously filtered if necessary, into a 1,500- pCi/ml. This solution is prepared from Na- ml beaker. If the sample contains more than tional Bureau of Standards’ encapsulated 350 mg/l of calcium, take a proportionately radium standard No. 4955 which contains smaller volume and dilute to 1,000 ml. If a 0.100 X curie of radium-226 in 5 ml of 5 sample is believed to contain sufficient quan- percent HNO,. Rubber gloves should be worn tities of barium, strontium, or other dis- in preparing a standard solution by the fol- solved or suspended material to add more lowing recommended procedure. than 1 or 2 mg to the weight of the pre- 5.5.1 Place the vial containing the cipitate, the internal standard modification radium standard in a clean, heavy-wall, should be used (see. 6.8). small-neck bottle or flask of 250- to 500-ml 6.2 Prepare duplicate standard solutions, capacity. Add 50 ml of 3 N HCl, and stopper each consisting of 5.00 ml of radium stand- securely with a polyethylene stopper. ard solution (1 ml=10 pCi radium-226) di- 5.5.2 Place the bottle (or flask) in a luted to 1,000 ml with distilled water, and strong plastic sack and, holding the stopper a blank solution consisting of 1,000 ml of firmly in place, shake vigorously to break distilled water. Add a few drops of methyl the vial. orange indicator, and adjust the pH to ap- 5.5.3 Decant the solution into a 2-liter proximately 3.5 by dropwise addition of con- volumetric flask. centrated HCl (or NH,OH followed by HCI 5.5.4 Rinse the bottle with 50 ml of if the sample has an initial pH below 3.5). 3 N HCl, and decant into the 2-liter flask. 6.3 Heat blank solution, samples, and 5.5.5 Add another 50 ml of 3 N HC1, and standard solutions to incipient boiling point, wash thoroughly using the ultrasonic and readjust the pH to approximately 3.5. cleaner. Decant into the 2-liter flask. 6.4 Add 3 ml of the barium carrier solu- 5.5.6 Rinse with 50 ml of 3 N HCl. De- tion to each beaker and stir vigorously. cant into the 2-liter flask. While stirring, add 15 ml of saturated am- 5.5.7 Repeat steps 5.5.4 and 5.5.5 al- monium sulfate solution. Continue heating ternately, three more times each. near the boiling point for an aidditional15 5.5.8 Dilute the solution in the 2-liter to 30 min. flask to 2,000 ml with distilled water and 6.5 Allow the precipitate to digest at mix thoroughly. room temperature for 4 hr or longer, then The final concentrations of radium and collect the precipitate on a membrane filter. hydrogen ion in this stock solution are: Police down the beaker carefully to ensure (Ra+2)=50pCi/ml and (Hf)=0.75 M. complete transfer of the precipitate to the 5.6 Radium standard solution 11,l ml= 10 filter. Wash the precipitate with small vol- pCi: Transfer 50 ml of the radium standard umes of the sulfuric acid wash solution. The stock solution to a 250-ml volumetric flask barium sulfate precipitate should be evenly METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 41

distributed to minimize self-absorption of ard procedure is calculated using the fol- alpha particles. A fine jet of the wash solu- lowing equation : -- tion can be used to redistribute the precipi- Ea=---,C,S--C, tate on the filter and to obtain uniform dis- d,, tribution. where - 6.6 When the membrane filter is practi- c,,=average count rate (cpm) of the calIy dry, mount it in the ring-and-disk sample containing the internal holders. standard, and 6.7 Determine the alpha activity of the d,, = known disintegration rate (dpm) blanks, standards, and samples after allow- of the internal radium standard ing the precipitates to age 15 d (or longer). added to the sample. Count each planchet for 100 min. 7.3 Calculation of radium gross alpha 6.8 For those samples suspected of hav- concentration, normal procedure. Use equa- ing excessive absorption loss of alpha radia- tion : tion, repeat the analysis using the internal 1000 c standard procedure as follows: Prepare a pCi/l of Ra (as radium-226) =- second sample as in 6.1. Add 5.00 ml of KVE,‘ radium standard solution (1 ml=10 pCi) . 7.4 Calculation of radium-gross alpha Proceed with the analytical determination as concentration, internal skandard procedure. before, and count. The difference between Use equation : the count rate of the sample containing the 1000 c pCi/l of Ra (as radium-226) =- internal standard and the count rate of the KvEa’ sample is the basis for calculating the effi- ciency factor to be used. 8. Report

1 7. Calculations Report concentrations less thgn 1.0 pCi/l to one significant figure and concentrations 7.1 The “radium as radium-226” effi- above 1.0 pCi/l to two significant figures. ciency factor (E,)is calculated using the fol- lowing equation : 9. Precision Reproducibility at the two standard de- viation level is approximately 51.0 pCi/l at where concentrations of 0.5 pCi/l and below. Re- - e,= average count rate of standard producibility is to approximately -1-20per- (cpm) corrected for background cent at higher concentrations. and blank, and d,= disintegration rate of standard Reference (dpm). Barker, F. B., and Johnson, J. O., 1964, Determina- 7.2 The “radium as radium-226” effi- tion of radium in water: U.S. Ged. Survey ciency factor (Ea),for the internal stand- Water-Supply Paper 1696-B, 29 p.

Rad i u m -226, d i ssol ved Radon emanation method (R-1141-76)

Parameter and code: Radium-226, dissolved (pCi/l): 0951 1

1. Application 3. Interferences The method is applicable to any water The method is normally specific for sample. radium-226. Radium-223 and radium-224 produce radon-219 and radon-220, respec- 2. Summary of method tively. Neither of these interfere directly, but the 10.6 hr lead-212 from radon-220has The method is based on the isolation of alpha-emitting daughters which could inter- radon-222 produced by radium-226 and fere. A wait of 2 or 3 d before counting measurement of the alpha activity of the eliminates the interference. The alpha-emit- radon and its short-lived daughters. The ting daughters of radon-219 have no effect method is specific for radium-226 in contrast if sufficient waiting time is allowed for com- to the precipitation method of Barker and plete decay of the 36 min lead-211. Johnson (1964). The procedure represents an improvement of the emanation method of 4. Apparatus Rushing (1967) in the substitution of a 4.1 Alpha-counting apparatus, scaler and complexing agent to redissolve precipitated high voltage power supply, preamp and am- barium sulfate. Formerly a complex proce- plifier with discriminator. dure for resolution involving a strong acid, 4.2 Beaker, 1,500 ml. ashing, and evaporation was required. Radon 4.3 Gas delivery system, for gas. is measured in a modification of the alpha 4.4 Mixer, wiggle-plate or ultrasonic type. scintillation cell of Lucas (1957). 4.5 Radon deemanation train and bubbler Dissolved radium in filtered water is col- (fig.5). lected by coprecipitation with barium sul- 4.6 Radon scintillation cell and housing fate. The precipitate is centrifuged and then (fig.6). dissolved in alkaline sodium diethylene tria- mine pentacetate solution. The solution is 5. Reagents transferred to a radon bubbler, and any 5.1 Barium carrier solution, 50 mg bar- radon present is removed by purging with ium/ml: Dissolve 75.81 g barium chloride helium gas. Fresh radon is then allowed to (BaCl,) in distilled water and dilute to 1,000 grow in. After several days the ingrown ml. radon is purged into an alpha scintillation 5.2 Defoaming emulsion, Dow Corning cell, short-lived daughters are allowed to Anti Foam H-10 emulsion, or equivalent: grow in, and the alpha-count rate is then Dilute to approximately 4 to 5 percent solu- determined. The radium-226 concentration tion with distilled water before using. in the original water sample is calculated 5.3 DPTA-TEA solution: Dissolve 10 g from the radon determination on the basis of sodium hydroxide pellets in a beaker con- of the rate of radon production with time. taining 60 ml of distilled water, and stir in

43 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS firmly in place, shake vigorously to break A - BUBBLER WITH POROUS DISK AND STIRRING BAR the vial. B - DRYING TUBE 5.4.3 Decant the solution into a 2-liter D C - SCINTILLATION CELL D - MANOMETER. aocm volumetric flask. CnE - ,HHLSI+IFMDELIVERY 5.4.4 Rinse the bottle with 50 ml of 3 N F - GEARED MOTOR WITH HCl and decant into the 2-liter flask. 11 $ ROTATING MAGNET 5.4.5 Add another 50 ml of 3 N HCl and wash thoroughly using the ultrasonic cleaner. Decant into the 2-liter flask. 5.4.6 Rinse with 50 ml of 3 N HCl. De- cant into the 2-liter flask. 5.4.7 Repeat steps 5.4.4 and 5.4.5 alter- nately, three more times each. 5.4.8 Dilute the solution in the %liter flask to 2 liters with distilled water and mix thoroughly. The final concentrations of radium and hydrogen ion in the stock solutions are: (Ra+2)=50 pCi/ml and (H+)=0.75 mole/l. 5.5 Radium standard solution 11, 1 ml= 1.000 pCi: Dilute 10.00 ml radium standard solution I and 10 ml of concentrated HCl to 500 ml with distilled water. 5.6 Sulfuric acid wash solution: Add 5 ml of concentrated HSO,and 3-5 drops of Tri- ton X-100 to 4 liters of distilled water. 5.7 Sulfuric acid, concentrated. Figure fi.-Radon deemanation vain and bubbler. 6. Procedure cold-water bath until dissolved. Add 20 g of 6.1 Coprecipitation of radium with bar- purified diethylene triamine penta acetic ium sulfate. acid (DPTA), and continiue stirring until 6.1.1 Add 5 ml concentrated hydro- dissolved. Add 17 ml of 50-percent triethano- chloric acid to 1,000ml of filtered water sam- lamine, mix and dilute to 100 ml. Store in ple contained in a 1,500-mlbeaker. Teflon bottle- 6.1.2 Add 1 ml of 50 mg/ml barium 5.4 Radium standard solution I, 1 ml= carrier to the sample and stir. 50.0 pCi :This solution is prepared from Na- 6.1.3 Cautiously add 20 ml of concen- tional Bureau of Standards’ encapsulated trated sulfuric acid to each sample with con- radium standard No. 4955 which contains stant stirring. (Use of a 500-ml dispensing 0.100X10-6 curie of radium-226 in 5 ml of flask fitted with a 50-ml delivery head facili- 5 percent HNO,. Rubber gloves should be tates the acid addition.) Stir well after the worn in preparing a standard solution by acid addition. Allow barium sulfate precipi- the following recommended procedure. tate to settle overnight. 5.4.1 Place the vial containing the 6.1.4 Carefully remove the supernate by radium standard in a clean, heavy-wall, decantation or suction, and quantitatively small-neck bottle or flask of 250- to 500-ml transfer the balance of the supernate and capacity. Add 50 ml of 3 N HC1 and stopper precipitate to a 40-mlcentrifuge tube using securely with a polyethylene stopper. a rubber policeman and small quantities of 5.4.2 Place the bottle (or flask) in a dilute sulfuric acid-Triton-X-100 wash solu- durable plastic sack, and,holding the stopper tion. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 45

G A- SCINTILLATOR CELL - KOVAR BELL COATED INSIDE WITH ACTIVATED PHOSPHOR B- QUARTZ WINDOW CEMENTED TO SCINTILLATOR BELL C- PHOTOMULTIPLIER TUBE D-TUBE BASE WITH H VOLTAGE DIVIDER E- POLYURETHANE CENTERING RING AND SUPPORT F- HIGH VOLTAGE AND SIGNAL CONNECTION G- LIGHT-TIGHT METAL HOUSING H-ELASTIC LIGHT TIGHT FELT SURROUND

Figure 6.-Radon scintillation cell and housing.

6.1.5 Centrifuge as necessary, decant, 6.2 Deemanations. and discard supernate. 6.2.1 Using a funnel with a fine tip, 6.1.6 Add approximately 10 ml of dis- transfer the cooled solution to a clean bub- tilled water and 1.5 ml of DTPA reagent to bler. Wash the centrifuge tube several times the precipitate in the centrifuge tube. Dis- with distilled water, and add the washings perse the precipitate in each tube by using a and sufficient additional water to the bubbler wiggle-plate mixer or an ultrasonic unit. to leave approximately 2 cm of airspace at Place tubes in a wire rack, and immerse rack the top. Add 1-3 drops of 4 percent silicone and tubes to a depth of approximately 1 inch defoaming emulsion to the solution in the in a boiling-water bath. bubbler to minimize frothing during purging. 6.1.7 Complete dissolution should occur 6.2.2 Attach stopcock and “0” ring to within a few minutes if the barium sulfate bubbler using clamp, leaving outlet stopcock “pellet” was adequately dispersed. Occasion- on bubbler assembly in open position. Attach ally, volume of solution in the centrifuge helium line (3-5 psi) to inlet side of bubbler. tubes may decrease by 4-5 ml as a result of Slowly olpenstopcock on inlet until a stream prolonged heating, and the precipitate may of fine bubbles rises from the porous disk. not dissolve. Addition of distilled water to Maintain a Isteady flow of bubbles through bring the volume to approximately 20 ml the sample for approximately 20 min to com- maximum plus additional redispersion and pletely purge all ingrown radon from the heating will usually result in rapid dissolu- solution. Close inlet stopcock and allow pres- tion of even difficulty soluble precipitates. sure under porous disk to equalize momen- After the precipitate has dissolved, cool the tarily. Close outlet and record the day, hour, tubes, and minute. This is zero time for the growth 46 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS of radon that will be removed in the second ing. Count overnight (1,000 min) for the deemanation and counted. average water sample. 6.2.3 Allow from 2 to 20 d in-growth 6.2.8 Dates, times, counts, and all other time for radon-222 depending upon the pertinent sample information should be re- radium-226 concentration in the original corded on data and calculation sheets. sample,volume of sample used, and so forth. 6.3 Calibration of equipment: One low 6.2.4 The second deemanation is made (~10cpm) and one high (~1,000cpm) by setting up the bubbler as in 6.2.2 except count rate disk standards are useful for rou- that both stopcocks are initially closed. At- tine instrument calibration tests and for de- tach bubbler to drying tube with “0” ring termining photomultiplier tube plateau and clamp. Evacuate purging assembly, in- curves. Prepare by precipitating each of two cluding cell, with vacuum pump for approxi- standards containing 5 and 500 pCi of mately 1.5 to 2 min. Close stopcock at vacuum radium-226with 50 mg of barium sulfate re- pump, turn pump off,and momentarily crack spectively as previously described. Mount the vacuum-pump connection. Open stopcock in precipitate by filtering through a 47-mm helium line above bubbler-inlet stopcock and 0.45-micrometer membrane filter. Dry and momentarily crack “0” ring connection to place on the disc of a ring-and-discassembly. purge trapped air from line and bubbler-inlet After drying, cover the precipitate on the fil- connection. Clamp and allow system to stand ter with a Mylar disk coated with an alpha- for approximately 2 min. If system leaks, sensitive pholsphomr.The dull phosphor-coated manometer meniscus will flatten or mano’me- side should be placed against the sample. ter will begin to fall. If meniscus remains Cover with the ring, press into place, and stable, proceed to next step. then seal the assembly with several pieces of 6.2.5 Carefully open bubbler-outlet stop- cellophane tape to prevent it frocm separating. The high count rate standard is used to de- cock until manometer begins to fall (check i porous disk for fine bubbles). Allow vacuum termine the plateaus for each photomulti- to equilibrate slowly (otherwise there is ex- plier tube, and the appropriate operating cessive risk of drawing liquid sample into voltage is then chosen accordingly. The low drying tube). Bubbling will slow appreciably count rate standard is used to check instru- in a few seconds. Slowly open outlet stopcock ment operating conditions at low count rates complete. Then continue with purging by comparable to those of typical samples. slowly opening bubbler-inlet stopcock, check- Frequently operating characteristics of ing porous disk carefully for rising bubbles. two or more photomultiplier-counting sys- (Flow rate must be closely controlled again tem are sufficiently similar to enable the use at this point, to prevent sudden surge of of a single high-voltage power supply. liquid into drying tube.) Allow pressure to Minor differences in the counting efficien- build up slowly, controlling manometer fall cy of each unit can be adjusted by the use of rate to complete purging in 15-20 min. To a focusing potentiometer on each photomulti- prevent cell leakage during counting, clwe plier housing. the cell stopcock at approximately 4 mm be- Long-term instrument backgrounds low atmospheric pressure. should be obtained for each counting system 6.2.6 Close down purging assembly stop- and should not generally exceed 0.005 cpm. cocks from cell to helium inlet in sequence as Scintillation cell background count rates rapidly as possible. Record time. Remove should be determined periodically for each bubbler from assembly quickly, and crack cell in combination with each instrument. outlet stopcock momentarily to release helium Generally, background count rates are deter- pressure. mined using a minimum of 1,000 min. Back- 6.2.7 Place cell in light-tight counting ground count rates for a specific cell may 3 chamber. Allow to age or 4 hr before count- vary considerably from one instrument to an- , METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 47 other, but should not generally exceed 0.10 to 7. Calculations 0.15 cpm. 7.1 Radon counting efficiency factors After long use or after counting a high- (E).The calculation requires corrections for radium-content sample, background rates in radon in-growthand radon decay. The radon some scintillation cells may become excessive in-growth and decay curves and their rela- (>OX cpm) for low-level work. In that tion to the time intervals that appear in the event the cells must be used only for rela- equation are shown in figure 1. Substitute tively high-level samples. Original low back- the experimental data obtained for each ground may be restored by rebuilding. cell-instrument unit (sec. 6.4) into a modi- fied form of equation 4. 6.4 Experimental determination of count- - ing efficiency :The counting efficiency of each E=- C9l , scintillation cell varies between cells and be- d, (1 -e- W)(e-124) tween counting instruments. Consequently, where for the most accurate wok,the counting effi- ciency of each scintillatioa cell should be de- x1 = decay constant of radon-222 termined in each instrument in which it is (1.259~10-~min-l), used. tl= time interval for buildup of radon between the previous The counting efficiency for each cell- deemanation of the standard instrument unit i,s determined by counting (point A,fig. 1) and the pres- radon transferred from a “standard bubbler” ent deemanation (point B), containing a measured amount of radium-226 t4= time interval between deemana- standard solution. A minimum of four or five tion standard (point B) and standards or one for each counting instru- midpoint of the counting ment enables four or five cells to be cali- t3 brated simultaneously. Waiting time for time,= tz+ -, 2 radon in-growth is also considerably reduced as compared to that required if only one where standard is available. t, = time interval between deemana- tion of the standard, (point Standards are prepared by pipetting 10.0 B) and the beginning of the ml of 10-pCi/ml radium-226 standard solu- count time (point C), tion directly into each of several bubbler t, = half the time interval between tubes. The tubes are fitted with an “0”ring the beginning (point C) and stopcock assembly, and then deemanated to the end (point D) of the determine the zero in-growth time for radon counting time, and - in the same manner as a sample. Barring any c, and d, are as defined in equation 2. spillage or breakage, the standards will last 7.2 Calculation of radium-226 concentra- indefinitely and can be deemanated every 4 or tion: Efficiency (E)used for an individual 5 d to provide radon for calibration purposeis. sample is that determined for the cell and Cell-counting efficiencies are generally instrument used to count the sample. An in- about 5.3 cpm/pCi of radon-222 after in- growth factor is introduced because of in- growth of daughters for 3 hr, but this may growth of radon with time after the first vary comsidwably depending upon factors deemanation. Use a modified form of equa- including the age of the cell, phototube condi- tion 3 when counting time is less than 3,600 tion, and moisture in cells. Erratic results minutes. are so’metimes obtained as a result of im- pCi/l of radium-226 proper cell or instrument grounding, loose lO0OC connections, noilsy power lines, and so forth. 48 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS where where J tl=time interval for in-growth of tz= delay before counting, point B to radon betwen first deemanation C,figure 1,and (step 6.2.1) and second deemana- t, = time interval of count, point C to D. tion (step 6.2.5) of the sample, and The other symbols are as defined in 7.1 t4= time interval between second de- and 7.2. emanation of the sample (step Radon decay and in-growth factors are 6.2.5) and midpoint of the sample easily and accurately calculated with sn elec- t, tronic calculator having natural log and e5 counting time, = tz +-. functions. The table of “Radon fraction 2 (e-xt)remaining after radioactive decay for The other symbols are as defined in sec- specified times,” for commonly used time in- tion 7.1. tervals (table 1) may also be used. When counting time exceeds 3,600 min- utes, use equation 3 including the term for correction for decay during the count. 8. Report pCi/l of radium-226 Report concentrations less than 0.10 1000 Zlt, pCi/l to one significant figure and values above 0.10 pCi/l to two significant figures.

Table 1.-Radon fraction (e-”) remaining after radioactive decay for specified times [Radon TH = 3.823 d] i Time Days Hours Minutes Time Minutes 0.834,18 0.992,47 0.999,87 0.996,lO .695,85 .985,00 .999,75 .995,98 .580,46 .977,59 .999,62 .995,85 .484,21 .970,23 .999,50 .995,73 .403,91 .962,93 .999,37 .995,60 .336,93 .955,68 .999,24 .995,48 .281,07 .9 48,4 9 .999,12 .995,35 .234,46 .941,35 .998,99 .995,23 .195,58 .934,27 .998,87 .995,10 .163,15 .927,24 .998,74 .994,98 .136,09 .920,26 .998,62 .994,85 .113,53 .913,33 .998,49 .994,73 .094,70 .906,46 .998,36 .994,60 .079,00 .899,64 .998,24 .994,48 .065,90 .892,87 .998,11 .994,35 .054,97 .886,15 .997,99 .994,22 .045,86 .879,48 .997,86 .9 94,lO .038,25 .872,86 .997,74 .993,97 .031,9 1 .866,29 .997,61 .993,8.5 .026,62 .859,77 .997,48 .993,72 .022,20 .853,30 .997,36 51 ______.993,60 .0 18,52 .846,88 .997,23 52 ______.993,47 .015,45 .840,50 .997,11 .993,35 24 _____ .012,89 .834,18 .996,98 .993,22 .010,75 ___--- .996,86 .993,10 .008,97 __-_-- .996,73 .992,97 .007,48 ______.996,61 .992,85 .006,24 __---_ .996,48 .992,72 .005,21 _____- .996,36 .992,60 .004,34 ______.996,23 .992,47 METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 49

, 9. Precision tion of radium in water: U.S. Geol. Survey On the basis of limited data the precision Water-Supply Paper 1696-B, 29 p. at the 0.10 pCi/l level is estimated at 220 Lucas, H. F., 1957, Improved low-level alpha scin- percent: Above 0.10 pCi/l the precision is tillation counter for radon: Rev. Sei. Instr., no. estimated at &10 percent 28, 680-683. Rushing, D. E., 1967, Determination of dissolved References radium in water: Am. Water Works Assoc. Barker, F. B., and Johnson, J. O., 1964, Determima- Jour. no. 59, 593-600. i Rad i u m -228, d issolved Determination by separation and counting of actinium- 228 (R-1142-76 )

Parameter and code: Radium-228, dissolved (pCi/l): none assigned

1. Application countered only in waters contaminated by The metho,d is applicable to all natural- reactor effluent. water samples. Applications to samples con- Success ob the method at low levels of taining reactor effluent or other contaminants radioactivity depends largely upon use of re- have not been evaluated. agents essentially free of radio sac t'ive con- tamination. Purity of the yttriutmreagent is 2. Summary of method especially important. The method is based on the chemical sepa- 4. Apparatus ration and subsequent beta counting of ac- 4.1 Centrifuge. tinium-228, the daughter of radium-228. 4.2 Centrifuge tubes,40- or 50-ml capaci- Radium-228 is not determined directly be- ty, heavy-walled . cause of the difficulty of counting its weak 4.3 Drying lamp, infrared, mounted in beta emislsion in the presence of other alpha- ringstand. emitting radium isotopes and their beta-emit- 4.4 Hotplate. ting daughters (Johnson, 1971). 4.5 Low-background counter, an antico- incidence-type counter with 2-in. thin win- 3. Interferences dow flowing gas propo,rtionaldetector prefer- No chemical interferences have been de- ably capable of measuring both alpha and tected. Because of chemical similarity, radio- beta activity simultaneously. nuclides of the elements and the 4.6 Planchets, 50-mm diameter, concen- rare-earth elements may accompany the ac- tric-rin,gtype. tinium precipitate. Significant concentrations 4.7 Stirring rods, Teflon. of these would be expected only in areas where nuclear fission or nuclear research ijs 5. Reagents carried on. 5.1 Ammonium hydroxide, concentrated. Radiochemical interferences are unlikely 5.2 Ammonium sulfate solution, 200 mg/ to occur in natural waters. Decontamination ml: Dissolve 200 g of ammonium sulfate in factors for other natural radionuclides ap- distilled water and dilute to 1 liter. pear to be 5,000 or greater. Exhaustive tests 5.3 Ammonium sulfidesolution, 2 percent : for artificially produced radionuclides have Dilute 10 ml of 20-23 percent aqueous am- not been made, but little or no interference monium sulfide solution to 100 ml. has been detected for cesium-137,strontium- 5.4 Barium carrier solution, 1 ml= 16.00 90, and yttrium-90. Lanthanum-140 appears mg BafZ: Dissolve 28.46 g 09 BaC1,.2Hz0 in to be the most probable interference,but this distilled water, add 2 ml of concentrated ni- and other rare-earth nuclides would be en- tric acid, and dilute to 1,000 ml.

61 52 TECHNIQUES OF WATER-RESOURCESINVESTIGATIONS

5.5 Binder solution: Dissolve about 1 g of trate solution containing 5-15 pg of barium. i “DUCO”cement in 100 ml acetone. Make up to volume with 1 N nitric acid. 5.6 Citric Acid, 1 M: Dissolve 210.1 g of Activity of radium-228 in the solution is citric acid monohydrate in distilled water calculated as follows : and dilute to 1,000 ml. Radium-228 (dpm/ml) 5.7 EDTA reagent, 0.25 M EDTA con- = 246W (1 -eo.1z05t) , taining 20 mg/ml NaOH: Dissolve 20 g of where NaOH in about 750 ml dis,tilled water. Heat, W = thorium concentration in mg/ml, and slowly add 93 g of disodium ethylenedi- and aminetetrascetate while stirring. When dis- t = years since thorium dioxide was solution is complete, cool, and dilute to 1,000 prepared. ml. 5.8 Lead carrier solution I, 1 ml=15.00 At equilibrium 1 mg thorium=246 dpm radi- mg Pb+”: Dissolve 23.97 g of Pb(NO,), in um-228. distilled water; add 2 ml of concentrated 5.15 Yttrium carrier solution, 1 ml= 18 nitric acid, and dilute to 1,000ml. mg Y+3:Add 22.85 g of Yzo3 to a 250-ml 5.9 Lead carrier solution 11, 1 ml=1.50 Erlenmeyer flask containing 20 ml of water. mg Pb+,: Dilute 100 ml of lead carrier solu- Swirl, place on a hotplate, and heat to boil- tion I to 1,000 ml. ing. Slowly and cautiously add concentrated 5.10 Methyl orange indicator solution. nitric acid, while stirring, until a homogene- 5.11 Nitric acid, concentrated. ous solution is obtained. (Usually about 30 5.12 Sodium hydroxide, 10 N: Dissolve mi of nitric acid is required. It is sometimes 400 g of NaOH pellets in distilled water and necetssary to add more water.) Add an addi- dilute to 1 liter. Store in a polyethylene or tional 70 ml of concentrated nitric acid and Teflon bottle. dilute to 1 liter. Note: The yttrium oxide 5.13 Sulfuric acid, 18 N: Cautiously add must be as free as possible of beta activity 500 ml concentrated H,SO, while stirring (usually due to ). Yttri- into about 400 ml distilled water; cool, and um oxide suitable for carrier preparation has dilute to 1,000 ml. been obtained from American Potash and 5.14 Radium-228 standard: The prepara- Chemical Corp.,West Chicago, Ill. tion of the standard solution is complicated 5.16 Yttrium-strontium carrier solution, by the fact that a standard of radium-228 is 1 ml=0.9 mg SrfZ and 0.9 mg Y+3:Dissolve not readily obtainable. Generally the radium- 434.8 mg of Sr (NO,) in distilled water ; add 228 must be obtained from thorium metal or 10 ml of yttrium carrier solution, and dilute pure compounds sufficiently old to permit to 200 ml. equilibrium in-growth of radium-228.In 40- year-old material the radium-228 daughter 6. Procedure has reached 99 percent of equilibrium con- 6.1 Use 1-4 liters of sample, on the basis centration. The following procedure is used of the expected radium-228 content, and to accurately prepare standardized thorium- evaporate to 1 liter, if necessary. Add 5 ml 232 and radium-228 from thorium oxide ap- of citric acid for each liter of original sam- proximately 50 yr old. ple, and then add a few drops of methyl 5.14.1 Weigh out a 1.0000 g sample of orange indicator. If the solution is yellow, thorium dioxide. add nitric acid until the red color is obtained. 5.14.2 Transfer the sample to a 250-ml Prepare standards by dilution of appro- Erlenmeyer flask containing 20 ml of acid priate volumes of the “old” thorium-radium mixture (15.7 N HNO,, 0.05 N HF). Care- solution to 1 liter. A disintegration rate of fully heat and stir the mixture until the Tho, approximately 1,000 dpm is appropriate. dissolves. Transfer the solution to a 200-ml Carry the standards through the procedure volumetric flask and add a little barium ni- in the same way as a sample. METHODS FOB DETERMINATION OF RADIOACTIVE SUBSTANCES 53

, 6.2 Add 10 ml of lead carrier solution I, 6.10 Transfer the aged solution to a cm- ' 2 ml of barium carrier solution, and 1 ml of trifuge tube, and add 0.3 ml of ammonium yttrium carrier solution. Heat to incipient sulfide solution. Add 10 N NaOH dropwise boiling, and maintain at this temperature for while stirring vigorously, until lead sulfide about 30 min with frequent stirring. precipitates ; then add about 10 drops excess. 6.3 Add concentrated ammonium hydrox- Stir intermittently over a period of about 10 ide until the yellow color of methyl orange is min. The lead sulfide should settle as a fine- obtained; then add a few drops excess. Pre- granular precipitate. cipitate lead and barium sulfates by adding 6.11 Centrifuge the solution, and transfer 18 N sulfuric acid until the red color reap- the supernatant liquid to clean centrifuge pears ; then add 0.25 ml excess. Add 15 ml of tubes. Discard the residue. ammonium sulfate solution for each liter of 6.12 Place the centrifuge tube in a hot- solution; keep the sample at a temperature water bath and slotwly add 10 N NaOH to the of about 90°C for 30 min and stir frequently. liquid while stirring, until yttrium hydroxide Remove from the hotplate and albw to settle precipitates; add 1 ml excess and stir inter- for at least 2 hr ; then siphon or decant most mittently for several minutes. Record the of the supernatant liquid and discard. date and time of the precipitation. Centrifuge 6.4 Transfer the precipitate to a 40-ml as soon as the yttrium hydroxide has largely centrifuge tube ; centrifuge, and discard the settleld. Dislcard the supernate. supernatant liquid. 6.13 Wash the precipitate thoroughly 6.5 Wash the precipitate twice with con- with 5 ml of water containing about 10 drops centrated nitric acid using centpif ugation of 10 N NaOH, cemtrifuge, and discard the wash techniques. Add 25 ml of EDTA re- wash solution. agent. Heat in a hot-water bath and stir in- 6.14 Transfer the preclipitate to a 5-cm termittently, adding a little additional 10 N counting planchet, using small volumes of \ NaOH if the precipitate does not dissolve distilled water, while drying under an infra- readily. red light lamp. The last washing should con- 6.6 Add 1 ml of strontium-yttrium car- tain 1 ml of binder solution. rier solution. Stir thoroughly; add a few 6.15 Count the sample in a low-back- drops of 10 N NaOH if any precipitate forms. ground beta counter for a sufficient length of 6.7 Add 1 ml of ammonium sulfate solu- time to obtain satisfactory counting statis- tion and stir thoroughly; then add glacial tics. Record date and time that count begins. acetic acid dropwise until barium sulfate pre- For most natural waters a counting period of cipitates, and add 2 ml excess. Allow the pre- about 300 to 500 min is desirable because of cipitate to digest in a hot-water bath until the low concentration of radium-228. the precipitate has largely settled. Centri- fuge,then discard the supernatant liquid. 7. Calculations 6.8 Add 20 ml of EDTA reagent, heat, 7.1 Radium-228 (actinium-228) efficiency and stir until the barium sulfate precipitate factor: The calculation requires a factor for dissolves. Repeat steps 6.6 amd 6.7. in-growth of the actinium daughter and for 6.9 Dissolve the barium sulfate precipi- decay in the time interval between separa- tate in 20 ml of EDTA reagent, and add 1 ml tion from the radium parent and beginning of yttrium carrier solution and 1 ml of lead of count. carrier solution. If any precipitate forms it The chemical recovery factor is included should be dissolved by adding a few drops of in overall efficiency and is not separately 10 N NaOH.Transfer the solution to a Teflon identified. Use equation 4 omitting the f,, or polyethylene container. Record the date texm. - and time. Age for 36 hr or longer. Cover the E= CIL sample to prevent evaporation. &(~-AL)(l-e-hltl) (e-AA) ' 54 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS where where ! A= decay constant of radium-228 t.,= decay time of radium-228 between (0.1205 yr-l), sampling and analysis in years, A, = decay constant of actinium-228 and (0.001885 min-*), t, = counting time in minutes. t, = elapsed time between certification of the standard and time of count in The other symbols are defined in section 7.1. yeus, t, = time interval between final purifica- 8. Report tion of barium-radium sulfate Report values of less than 1.0 pCi/l to one precipitate (step 6.9) and separa- significant figure and values above 1.0 pCi/l tion of actinium daughter (aging to two significant figures. time in minutes) (step 6.12), and t, = time interval between separation of 9. Precision actinium daughter (step 6.12) and countin,g in minutes. The limited data available suggest that precision of the method is approximately 7.2 Calculation of radium-228 concentra- t20 percent for a concentration of 1 pCi/l tion: Use equation 3 omitting the f term. and +lo percent at the 10 pCi/l level or Correction for decay of radium-228is seldom above. required. In addition to in-growth and decay times for actinium-228 cited under 7.1 a cor- rection for decay during counting is required. Reference

pCi/l of radium-228 Johnson, J. O., 1971, Determination of radium-228 in - 1,oooc Alt, natural water: U.S. Geol. Survey Water-Supply KvE(e-hts)(1-e-hltl) (e-hlt2) (~-e-~ta)’ Paper 1696-G, 26 p. Radioruthenium, dissolved, as ruthenium-106 Distillation method (R-1150-76)

Parameter and code: Radioruthenium as ruthenium-106, dissolved (pCi/l): none assigned

1. Application enium to the tetroxide and to provide a high This method is applicable to analysis of boiling point for the solution. Ruthenium nonsaline waters which do not contain high tetroxide is then distilled, with the aid of air concentrations of organic matter. Using a bubbles through the solution, into a cold maximum sample volume of 100 ml, the de- sodium hydroxide solution. Ethyl alcohol is tection limit is approximately 4 pCi/l. Be- added to precipitate the ruthenium as mixed cause of this rather high detection limit, the oxides. method is of value primarily for water sam- The ruthenium oxides are separated from ples which have substantial fission product the supernate by centrifugation, redissolved activity. The sensitivity may be improved in hydrochloric acid, and the by preconcentrating the ruthenium and by ruthenium metal-oxide mixture by addition counting the sample more than the recom- of turnings. After dissolving the mended 150 min. excess magnesium, ruthenium is recovered by filtration through a tared filter and 2. Summary of method counted on a low-background beta counter. Recovery is determined gravimetrically. The Ruthenium salts form the tetroxide in corrected activity is compared that of acid solution when a strong oxidizing agent with is present. This oxide melts near room tem- standards similarly prepared to determine perature and can be easily volatilized from the ruthenium-106 activity of the sample. concentrated perchloric acid solution, which The restriction of the method to fresh- boils at approximately 200°C. Distillation of waters and low-sample volumes is necessary the tetroxide provides an excellent separa- to avoid large amounts of solids in the dis- tion from almost all potential sources of in- tillation flask. The range of usable sample terf erence. volumes as a function of dissolved solids The method used is essentially that de- has not be evaluated. scribed by Glendenin (1951). The water sample is evaporated to small volume, and 3. interferences stable ruthenium carrier is added. Nitric Because the analytical method separates acid is added to oxidize organic matter, phos- pure ruthenium,only ruthenium isotopes can phoric acid is added to prevent volatilization interfere. On the basis of half-life, ruth- of molybdenum, bromide and iodide are enium-103 is the only interference of added as holdback carriers, and so- practical significance. The ruthenium-103 dium bismuthate is added to oxidize the contribution to the total count is less than halide ions to their respective nonvolatile ruthenium-106contribution (for equal num- oxyacids (Wyatt and Rickard, 1961). Final- bers of atoms) because of lower counting ly, perchloric acid is added to oxidize ruth- efficiency. The ruthenium-103 contribution

66 56 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS decreases with time and may be completely TO LOW PRESSURE eliminated by several months decay. Alter- natively, the relatively low-energy betas /AIR SUPPLY from ruthenium-103 may be eliminated by counting the sample through an absorber which does not: significantly affectthe higher energy betas, from rhodium-106 (ruthenium- 106). Any organic matter present in the sample should be destroyed by boiling with a con- centrated nitric-perchloric acid mixture be- fore proceeding with the perchloric acid dis- tillation. To prevent explosive decomposition of the distillation mixture, nitric acid should be present as long as any unoxidized organic material remains. Large amounts of insoluble matter in the distillation flask may cause serious bumping E and impair the recovery of ruthenium. B 4. Apparatus lol 4.1 Compressed air, 1-2 psi supply for bubbling air through distillation flask. 4.2 Low-background counter, an anti-co- GROUND incidence-type counter with 2 in. thin win- A - FLASK WITH dow flowing gas proportional detector pre- GLASS JOINT (125ml) ferably capable of measuring both alpha and B - HOTPLATE beta activity simultaneously. 4.3 Burner, Bunsen, small. C - ANNULAR LEAD WEIGHT 4.4 Centrifuge, capable of accepting 40- ml tubes, D - HEAVY WALL PYREX 4.5 Centrifuge tubes,40-ml, heavy-wall. CENTRIFUGE TUBE (40ml) 4.6 Distillation apparatus. See figure 7. 4.7 Filter disks, Versapore, 47 mm, 5 E - ICE BATH micrometer (Gelman Instrument Co., Ann Figure -/.-Apparatus for distillation of ruthenium Arbor, Mich. 48106). tetroxide. 4.8 Hotplate. 4.9 Vacuum desiccator. 5.5 Magnesium metal turnings. 4.10 Vacuum-filtration apparatus for 47- 5.6 Nitric acid, concentrated. mm filters. 5.7. Perchloric acid, 70 percent. 5.8 Phosphoric acid, concentrated (85 5. Reagents percent). 5.9 Ruthenium carrier solution: Suspend solution, 1 10 mg 5.1 Bromide-iodide ml= 5 g of purified RuC1, in approximately 250 Br and 10 mg I :Dissolve 4.7 g NaI and 5.2 g rnl of 6 M HC1 and agitate for several hours NaBr in 400 ml distilled water. or overnight on a mechanical shaker. Filter 5.2 Diethyl ether. the resulting solution through a fine filter. 5.3 Ethanol, 95 percent. The solution is standardized by direct reduc- 5.4 Hydrochloric acid, concentrated and tion of aliquots with magnesium metal, fol- 6M. lowed by filtration. and weighing of the METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 57 mixed ruthenium oxide and metal produced. ruthenium. Note: It is best to perform the Care should be exercised to dissolve excess distillation behind an explosion shield. magnesium with HC1 and to thoroughly 6.7 Continue the distillation until the wash the precipitate before drying. dense white fumes of perchloric acid have 5.10 Ruthenium standard: Prepare stand- been carried over for a few minutes. Total ard deposits of ruthenium-106on filter paper time required for the distillation is usually starting with ruthenium standard solution 20-30 min. of 100 pCi in 5 ml. Follow steps 6.12 through 6.8 Lower the centrifuge tube from the 6.16 in the “Procedure” section. Correct the delivery tube, and remove the still from the activity determined for decay. hotplat,e.Disconnect the air supply and al- 5.11 Sodium bismuthate. low the still to cool. When cool, rinse the 5.12 Sodium hydroxide, 6 M. contents of the delivery tube into the centri- fuge tube with a little distilled water from 6. Procedure a washbottle. 6.9 Carefully warm the centrifuge tube to 6.1 Estimate the volume of sample on the near boiling over a small burner. Remove basis of dissolved solids. The volume is se- from the flame and add, in small amonnts, 3 lected to avoid excess precipitation of salts ml of 95 percent ethanol, reheating to in- in the evaporation step. The distillation flask cipient boiling after each small addition. It is holds 50 ml without danger of losing sample. important to swirl the solution constantly Sample volumes in excess of this are evapo- while heating. rated to 50 ml in a Teflon evaporating dish. 6.10 Cool the solution, and centrifuge to 6.2 Transfer the water sample into the recover the precipitated ruthenium oxidee. distillation flask, and evaporate down to ap- The supemate should be colorless and may proximably 5 ml. Add an accurately meas- be discarded. J ured amount of ruthenium carrier between 6.11 Suspend the precipitate in 10 ml of 20 and 30 mg. Record pipet volume to 0.01 distilled water to which 1 ml of 6 M NaOH ml. has been added. Heat (with swirling) to boil- 6.3 Evaporate to less than 5 mi. (Pre- ing. Centrifuge and discard the supernate. cipitation will usually occur, but the residue 6.12 Suspend the precipitate in a few mil- must not go dry.) Cool the flask and ’con- liliters of 6 M HCl and reheat to boiling tents. (swirl vigorously). Continue heating for a 6.4 Add the following reagents in order: few minutes to dissolve all solid material. Set 1 ml concentrated nitric acid, 1 ml concen- aside to cool. trated phosphoric acid, 2 ml bromide-iodide 6.13 Add about 5 ml distilled water to the reagent, 0.5 g sodium bismuthate, and 10 ml ruthenium solution ; then add small portions 70 percent perchloric acid. Mix thoroughly. of magnesium turnings to reduce the ruthen- 6.5 Place the lead-ring weight around the ium to the mixed oxide-metal precipitate. The flask. Moisten the ground-glass joint with solution may have to be heated to boiling as phosphoric acid, and attach the delivery ap- the reaction nears completion (supernate has paratus to the flask. Place the distillation light-blue color) to speed the process. apparatus on a hotplate, and immerse the 6.14 Add (slowly, at firlst) about 5 ml of end of the delivery tube in a 40-mlcentrifuge concentrated HCl, and then heat to boiling tube containing 12 ml of 6 M NaOH. The for a few minutes to dissolve any excess centrifuge tube is kept immersed in an ice magnesium. bath. 6.15 Filter the suspension through a tared 6.6 Connect the air supply to the still, 47-mm Versapore filter wing a vacuum-fil- and adjust the flow rate to a few bubbles per tration apparatus. (The Versapore filter second. Heat the flask slowly to the boiling should be washed and dried by the same pro- point, and then heat strongly to distill the cedure ais the sample (see below) prior to 58 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS taring). Wash the precipitate with three 10- 7.2 Calculation of ruthenium-106 concen- to 30-ml portions of boiling water, two or tration: Use equation 1. The ruthenium re- three 10-ml portions of 95 percent ethanol, covery factor (f) .is determined from the and with three 10-ml portions of diethyl initial and final weights of ruthenium carrier. ether to dry and distribute the precipitate pCi/l of ruthenium as ruthenium-106 over the filter pad. 10ooz 6.16 Remove the dry filter and place in a - (1) heated (100°C) vacuum desiccator for 5-10 KVEf(@-At)' min (no longer). Remove, cool in a desic- cator, and weigh. Mount in a ring-and-disk 8. Report assembly, covering the precipitate with a thin Report ruthenium-106 values o,fless than plastic kitchen-type map. 10 pCi/l to one significant figure. Values 6.17 Count the sample for three 50-min greater than 10 pCi/l are reported to two periods in a low-background beta counter. significant figures. 7. Calculations 9. Precision 7.1 Ruthenium-106 efficiency factor (E): Counting efficiency for ruthenium-106 is de- Typical ruthenium-106values fo,rthe pre- termined b~ measurement of the count rate cision of this method are +-4pCi/l or k20 of standards prepared by direct precipitation percent, whichever is larger, at the 95-per- of ruthenium metal from aliquots of the cent confidencelevel. ruthenium-carrier solution spiked with known amounts of wthenium-106. Rutheni- References um is reduced by metallic magnesium, fil- tered, dried, weighed, and counted a.~ in the Glendenin, L. E., 1951, Improved determination of procedure above. Since recovery is usually in ruthenium activity in fission, in Coryell, C. D., excess of 90 percent for the entire analytical and Sugarman, N., Radiochemical studies : The procedure, the counting efficiency on the di- fission products (Book 3) :New York, McGraw- Hill Book Co., Paper no. 260, p. 1549. rectly prepared standards is a reasonably Wyatt, E. I., and Richard, R. R., 1961, The radio- accurate measurement of the counting effi- chemistry of ruthenium: Natl. Aead. Sci., Natl. ciency of the samples. A typical value is 1.0 Research Council Nucl'ear Sci. Ser. NAS-NS cpm/pCi. 3029, 78 p. St ron t i u m'-90,d i ssolve d Chemical separation and precipitation method (R-1160-76)

Parameter and code: Strontium-90, dissolved (pCi/l): 13503

1. Application the analysis is improved by counting after The method is applicable to all natural 21 d to allow in-growthof yttrium-90. freshwaters and saltwaters. It is applicable If strontium-89 is present, it is counted as to reactor wastes and may be applicable to strontium-90. Approximately 45 percent of industrial wastes, provided that recovery the original strontium-89 activity is retained tests are made to assure that organic matter 21 d after the initial precipitation. Stronti- does not hold back strontium. um-90 may be determined independently by chemically separating the yttrium-90 daugh- 2. Summary of method ter, and relating its activity back to stronti- um-90. The composition of a mixture of the This method is based on the wolrk of Hahn three isotopes, strontium-90, strontium-89, and Straub (1955) and of Johnson and and yttrium-90,may be determined approxi- Edwards (1967). Dissolved radiostrontium mately by plotting the growth of radioactivi- is determined by beta counting after a ty with time as shown in Johnson and Ed- lengthy separation procedure that removes wards (1967). other fission products. Stable strontium car- rier is added to the sample, and a car- 3. Interferences bonate precipitation is made to collect strontium-90 accolmpanied by some fission Interferences from both fission products products. The carbonate preoipitate is dis- and natural radioactivity are negligible solved, and strontium nitrate mixed with cal- (Glendenin, 1951). As indicated above, stron- cium nitrate is precipitated with fuming tium-89 is counted as strontium-90. nitric acid. Calcium is removed by washing with acetone, and strontium nitrate is fur- 4. Apparatus ther purified by solution and reprecipitation. 4.1 Filter disks, 25-mm Whatman No. 42, The precipitate may still contain traces of or equivalent. fissio'n protducts such as transition metals) rare earths, niobium,zirconium, yttrium, and 4.2 Filtration apparatus for 25-mm mem- bar?ium. Radium may be present. These are brane filters. dl removed by the addition of iron and bari- 4.3 Ho tpllute. um carriers which are precipitated as the hy- 4.4 Low-background counter, an anticoin- droxide and chromate respectively. cidence-type counter with 2-in.thin window Final purification is effected by precipitat- flowing gas proportional detector prefer- ing strontium as the oxalate. Yttrium-90,the ably capable of measuring both alpha and radioactive daughter of strontium-90, has beta activity simultaneously. been completely removed at this point, and a 4.5 Ring-and-disk mounting assemblies new equilibrium begins. The yttrium isotope for filters, including 21-mm diameter rigid is also a beta emitter, and the sensitivity of copper backing disks.

59 60 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

5. Reagents tration of 0.1 to 10 mg/l. Strontium-90 5.1 Acetic acid, 6 M. standard activity (A),after a decay time 5.2 Acetone, anhydrous. (tn), for an initial activity (Ao),can be 5.3 Aluminum foil, 3.5 mg/cm2 or less. calculated by use of the following equation : 5.4 Ammonium acetate, 6 M solution. A =Aoe-htn, 5.5 Ammonium hydroxide, concentrated where and 6 M. hn = decay constant of strontium-90 5.6 Ammonium oxalate, saturated solu- (1.999 x months-*), and tion. t, = elapsed time (in months) bekween 5.7 Barium carrier solution, 1 ml= 3.00 certification of standard and time mg Ba+2: Dissolve 1.33 g of BaCl,-2H2Qin of count. distilled water. Add a few drops of con- centraked nitric acid and dilute to 250 ml. 5.17 Thymolphthalein indicator solution. 5.8 Diethyl ether. 5.9 Ethanol, 95 percent. 6. Procedure 5.10 Iron carrier solution, 1 ml=5 mg A reagent blank should be run with each Fe+3: Dissolve 500 mg of pure iron wire in set of samples to check for contamination a slight excess of nitric acid, and dilute to of reagen,ts,and to permit an accurate blank 100 ml. correction to be made. Occasional spiked 5.11 Nitric acid, fuming, concentrated, 6 samples (samples containing a known M and 1 M. amount of added standard) should also be 5.12 Phenolphthalein indicator solution. run through the entire procedure as checks. 5.13 Sodium carbonate solution, 1 M and 6.1 To a 1,000-ml or other suitable ali- 0.1 M. quot of filtered water sample in a 1,500-ml 5.14 Sodium chromate solution, 1.5 M. beaker add 20 mg of strontium carrier (5 5.15 Strontium carrier solution, 1 ml= ml of 4-mg Sr+2/ml). Heat to boiling on a 4.00 mg Sr+2: Dry “anhydrous” strontium hotplate. Make the solution basic to thymol- nitrate overnight at 105”C, and cool in a phthalein indicator (blue color) by drop- desiccator. Weigh out 9.66 g of the dried wise additions of concentrated ammonium salt, and dissolve in distilled water. Add 2 hydroxide, and add an additional 6 ml. ml of concentrated nitric acid, and dilute to 6.2 Add 15 ml of 1.0 M sodium carbon- 1 liter. A more exacting procedure is des- ate. Stir thoroughly, cover the beaker with cribed by Glendenin (1951, p. 1461), but the a watch glass, and digest on a steam bath for added work is not warranted unless the 1 hr. Add more ammonium hydroxide, if re- strontium nitrate contains radioactive im- quired, to maintain the blue color of thymol- purities which contribute significantly to the phthalein. (Add more indicator if the color reagent blank. fades.) 5.16 Strontium-90 standard solution: 6.3 Remove the beaker from the steam Strontium-90 standard solutions calibrated bath, and allow the precipitate to settle while to 2 1.5 percent are commercially available. the sohtion cools to room temperature. In purchasing standards it is essential that 6.4 Carefully decant or draw off as much the concentration of the stable isotopic car- as possible of the supernatant solution with- rier be known. Dilute the standard to ap- out disturbing the precipitate. Transfer the proximately 5 pCi/ml as described by Barker insoluble material to a 50-rnl Pyrex centri- and Robinson (1963,p. A28). It is generally fuge tube, and police the beaker with 0.1 M necessary to add both acid and inactive stron- sodium carbonate. Add washings to the cen- tium carrier at the time of dilution. The trifuge tube. Centrifuge and discard the final solution should be approximately 0.1 N Supernatant liquid. in hydrogen ion (HCl or HNQ,)and should 6.5 Cautiously add 1 N nitric acid drop- have a chemical strontium carrier conan- wise until the carbonate precipitate is com- METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 61 pletely dissolved. Dilute to 5 ml with dis- and stir. Allow the precipitate to settle and tilled water. (Note: There may be a small the solution to cool to room temperature. residue of silica or insoluble metal oxide re- 6.14 Filter through a tared 25-mm filter maining.) Very cautiously add 25 ml of fum- paper (Whatman, No. 42) supported on a ing nitric acid (work in the hood, use plas- sintered glass disk of a microanalytical fil- tic gloves, safety goggles or face shield, and tration assembly. Wash the precipitate with avoid inhaling the vapors!). Cool, stir, and three 5-ml portions of distilled water and centrifuge. then successively with small volumes of 95 6.6 Using the above safety precautions, percent ethanol and diethyl ether. Dry in a pour off the nitric acid as completely as pos- desiccator. Weigh as strontium oxalate sible, cool the tube below room temperature monohydrate, and determine the gravimetric using cold water or ice bath, and cautiously yield. add 25 ml of anhydrous acetone to the resi- 6.15 Mount the filter on a copper disk, due. Stir thoroughly and centrifuge, again 21-mm diameter. Cover with aluminum foil discarding the supernate. (3.5 mg/cm2 or less). Store the sample for 6.7 Repeat step 6.6. 21 d to permit in-growth of yttrium-90 to 6.8 Dissolve the nitrates in 5 ml of dis- 99.5 percent of equilibrium activity. If it is tilled water, and place the tube in a boiling- desired to break down the activity into the water bath until the odor of acetone is gone. three radioactive isotopes, it is necessary to (Caution: be sure acetone is completely re- start the counting within a few days of step moved.) 6.15 so that the decay of strontium-89 and 6.9 Cool below room temperature, and in-growth of yttrium-90 may be observed. again precipitate strontium nitrate by add- 6.16 Count the sample for 100 min on a ing 25 ml of fuming nitric acid. Swirl, cool, low-background, anticoincidence shielded and centrifuge. Discard the supernate. beta connter. 6.10 Dissolve the precipitate in 10 ml of distilled water and add 5 mg of ferric carrier 7. Calculations (1 ml of 5-mg Fe+3 per ml) and 15 mg of barium carrier (5 ml of 3-mg Ba+2 per ml) . 7.1 Strontium-90 efficiency factor (E) Add concentrated ammonium hydroxide with and fractional chemical recovery (fn) : Use constant stirring until ferric hydroxide be- equation 2. Determine fn from the weight of gins to precipitate. Then add several drops strontium oxalate. excess. Centrifuge, and decant the supernate - into a clean centrifuge tube. This operation removes the yttrium daughter of strontium- 90. Record the time and date. where 6.11 Add phenolphthalein indicator, and A= decay constant of strontium-90 add 6 N nitric acid dropwise until the pink (1.999 X month-I) . color disappears. Add 1 ml of 6 N acetate 7.2 Calculation of strontium-90 concentra- acid and 2 ml of 6 M ammonium acetate. tion: Use equation 1, making the decay cor- Place the samples in a boiling-water bath rection if necessary. Determine f from the and add 1 ml of 1.5 M sodium chromate weight of strontium oxalate precipitate. while agitating. Leave in the water bath for 1,000 c 5-10 minutes. pCi/l of strontium-90= (1) 6.12 Cool to room temperature, centri- KVEf (e-kt) fuge, and decant the supernate into a 100-ml According to Hahn and Straub (1955), beaker. Discard the precipitate. the chemical recoveries range from approxi- 6.13 Add 2 ml of concentrated ammon- mately 72 to 80 percent. Accurate determi- ium hydroxide, and heat to boiling. Add 5 nation of chemical recovery requires knowl- ml of saturated ammonium oxalate solution edge of the natural strontium content of the 62 TECHNIQUES OF WATER-RESOURCESINVESTIGATIONS sample. Normally this is less than 1 mg/l 9. Precision I and may be neglected. In a few cases, par- Minimum detection level is 0.5 pCi/l and ticularly with waters from an area extending this is also the precision at activities below through parts of Wisconsin and Illinois, the 5 pCi/l. At higher strontium-90 activities strontium content may equal or exceed the the precision is s 10 percent. strontium carrier added, and natural stron- tium must be determined for the calcula- tion of chemical recovery. References 7.3 Calculation of strontium-90, yttrium- 90 and strontium-89. The count rate is de- Barker, F. B., and Robinson, B. P., 1963, Deter- termined at intervals over a timespan of ap- mination of beta activity in water, U.S. Geol. proximately 30 d starting as close to zero Survey Water-Supply Paper 1696-A. time as possible. A typical counting sched- ule thereafter would be at 24 hr, 48 hr, 100 Glendenin, L. E., 1951, Determination of stronbium and barinm activities in fission, in Corydl, C. D., hr, 200 hr, and 400 hr. The count rates are and Sugarman, N.,eds., Radiochemical studies : plotted against time, and the curve thus ob- The fission products, National Nuclear Energy tained is compared against type curves for Series IV, 9, Paper no. 236: New York, different isotopic ratios of strontium-90 to McGraw-Hill,p. 1460-69. strontium-89 (Johnson and Edwards, 1967) Hahn, R. Z.,and Straub, C. P.,1955, Determination to obtain the ratio for the sample. of radioactive strontium and barium in water: Am. Water Works Assoc. Jour., v. 47, p. 335- 8. Report 340. Report strontium-90 equivalent activities Johnson, J. O., and Edwards, K. W., 1967, Deter- to S0.5 pCi/l or to &lo percent, whichever mination of strontium-90 in water: U.S. Geol. is greater. Survey Water-Supply Paper 1696-E, 10 p. Tritium Liquid scintillation method, Denver lab (R-1171-76)

Parameters and codes: Tritium dissolved (pCi/l): 09005 Tritium, in water molecules (Tu): 07012

1. Applicafion secondary scintillator used to shift the The technique is generally applicable to wavelength of the scintillations to the most the determination of tritium artificially in- sensitive spectral region of the PM tube. troduced into water by such activities as Much of the tritium data reported in the tracer experiments, nucleax power, waste literature was determined by use of the above disposal, and thermonuclear weapom testing. scintillation mixture. The dioxane-PPO com- The technique is not sufficiently sensitive to bination has been superseded by proprietory be applicable to the determination of very scintillators that produce gels when mixed low natural tritium levels. with water in proper ratios. The newer scin- tillators have approximately doubled the 2. Summary of method sensitivity of liquid scintillation counting for tritium. Since the composition of the pro- Liquid scintillation counting is based on prietory mixtures is not available, these are the conversion of the energy of a particle listed by trade name in the "Reagent" section. emitted by a radioactive nucleus to light ener- gy by means of a scintillating chemical. The While the mechanism of liquid scintillation scintillations are detected by a photomulti- counting is not completely understood, it plier (PM) tube. The electrical signals from seems clear that the energy transfer is a two- the PM tube are amplified and sent through stage process, with initial energy transfer to a simple multichannel analyzer (three or the solvent followed by transfer from the four channels at most) where sorting into solvent to the scintillator. Many substances, energy takes place. The counts in each chan- including water, interfere in the energy nei are displayed on a scaler and may be read transfer process to quench the sointillations out on paper tape, punched tape, or magnetic and reduce the count rate. Excessive salt con- tape. tent, certain metals, and organic compounds Liquid scintillation counting is primarily quench in varying degree. Colored substances used for the counting of beta emitters d- may quench by light absorption in addition though it can also be used for alpha-emitting to interfering with energy transfer. Quench- isdopes. ing substances are generally removed by When liquid scintillation counting is used vacuum distillation. to determine a radionuclide in aqueous sdu- The que'nching effect of water is compen- tion, the water sample is dissolved or dis- sated by using a constant volume of water persed in a large volume of organic solvent and constant ratio of water sample to scin- containing the scintillating chemical. A wide- tillator mixture in both samples and stand- ly used mixture for aqueous solutions was ards. The count rate of a particular sample dioxane-naphthalene containing the scintil- depends on both the volume of sample and lator 2,5-diphenyloxazole (PPO) and a the ratio of liquid scintillator mixture to

63 64 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS water sample. As the fraction of water sam- Operation-pTogrammable and automatic. I ple in the water scintillator mixture in- Internal check-an externd standard and ratio computation capability required. creases, count rakes increase until a point is Readout-automatic printout. reached where the quenching effect of the additional tritiated water exceeds the effect 4.3 Pipets, 8 ml. of the increased radioactivity. It is found 4.4 Vacuum-distillation apparatus. Con- that a practical compromise between maxi- sists of a 50-ml round-bottom flask as the mum sample volume and scintillator volume distillation flask and a 125-ml round-bottom provides optimum sensitivity. The maximum flask as the condenser flask. The distillation on the curve of activity versus volume of flask is heated with a rheostat-controlled tritiated water (in a constant overall vol- mantle, and the condenser flask dips into a ume) is a rounded plateau. Therefore, the Dewar containing isopropanol-dry ice. The proportion of water sample to scintillating two flasks are connected by a 20-mmdiame- liquid mixture is not critical and is easily ter U-tube,10-em long with ground-glass con- reproducible in routine work. nectors and stopcock for application of The permissible ratio of water sample to vacuum. Heating tape is coiled around the liquid scintillator mix is also controlled by U-tube connecting the two flasks. physical stability of the gel formed. 5. Reagents A detailed description of the method is given by Schroder (1971). 5.1 Scintillator, Instagel (Packard In- strument Co.) for low-temperature counting 3. Interferences (1”-4”C). A preblended gel-forming Distillation is used to remo,veboth quench- scintillator designated 3A70” (Research ing substances and radionuclides that could Products International) for room-tempera- contribute to excess count. Distillation is ture counting. fully effective in removing inorganic salts 5.2 Tritium standard solution: Appropri- ’! and high-boiling organic compounds. Or- ate standards are prepared by the dilution of ganic materials that vaporize at a lower tem- NBS standard tritiated water with “dead” perature than water and are condensable water, that is, water containing less than 1 under the same conditions as water vapor Tu. may be carried over in the vacuum distilla- 5.3 Water, “dead.” The tritium blank in- tion. If these materials contribute either troduced by reagents must be tested at inter- quenching or beta radioactivity the possibili- vals by analyzing “dead” water (water with ty of interference exists, although it would no measurable tritium content) in exactly appear remote. the same procedure as for a normal water Further protection against interferences sample. It is very difficult to confirm that a is provided by the energy discrimination in water is truly “dead,” and it is usually neces- the liquid scintillator analyzer and the ex- sary to make the assumption that water from ternal standard-ratio test. a deep well in a confined aquifer a hundred 4. Apparatus kiIometers or more from the recharge area is 4.1 Autopipet, 25-ml maximum capacity. “dead.” This assumption can be correct only 4.2 Counting equipment. Liquid scintilla- if the well is pumped sufficiently to expel all tion spectrometer ; counting systems capable meteoric water that may have entered the of meeting the following specifications : well and surrounding aquifer by leakage. Counting effic’iency-not less than 24 percent with optimum sample-scintillator mixture and 6. Procedure polyethylene vials. 6.1 Counting. Background-not to exceed 5 cpm at sea level . in tritium channel. 6.1.1 Distill 25 ml of water sample for Sample capacity-at least 100 samples. direct counting, or approximately 10 ml of METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 65

I , electrolyzed water sample, using the vacuum end of the counting run (see instrument in- j distillation apparatus. The U-tube is pre- struction manual). This procedure is a check heated to assure dryness. for quenching. The ratio of low-energy 6.1.2 Pipet 8.00 ml of the distillate from counts in channel A of the spectrometer is I I the preceding step into a 25-m1 polyethylene established for both tritium and an external vial, and add 14 ml of scintillator mix. The standard Placed near the !Sample vial. The choice of scintillator depends on the type of Value Of R in the following equation should liquid scintillation spectrometer to be used be constant to within &0.2. for counting. Instagel is used with the in- Counts in A , tritium] struments that operate at cold temperature .=[ Counts in B (3"C), and 3A70" is used with the instru- Counts in B standard.]. ments that operate at room temperature. Cap [Counts in A the vial and mix. Heat the Instagel-samPle Individual samples that fall outside this mixture on a hotplate at approximately range must be reanalyzed. If all values of R 100°C for 2 or 3 min. This clarifies the mix- fall outside this range, the ecintillator has ture. The 3A70*-samplemixture does not re- deteriorated and must be replaced. quire heating. The above operation is carried out under subdued red light to filter out the 7. Ca~cula+ions blue region of the spectrum. This minimizes excitation of fluorescence background in the Several statistical schemes have been pre- sample. sated for the calculation and verification of 6.1.3 Prepare three blanks and two tritium data, each intended to optimize a par- standards in the same manner as the sam- ticular analytical situation. In the present ples. Place one standard in the 2d counting analytical procedure the repeated counting effect ) position in the Spectrometer and one in the of individual samples has the of aver- 10th position. Place blanks at intervals aging out short-term fluctuations. Statistical throughout the run of 10-14 samples. checks have shown that highest precision,in 6.1.4 Place three sealed standards (tri- this analytical situation, is attained by the ated toluene in glass-sealed scintillator solu- use of longer term average values for the tion) in the . One standard goes in the samples and standards. first counting position to permit monitoring Counting efficiency and background values the instrument before counting the samples. are determined with tritium standards when- The remaining standards are placed ever a new lot of scintillator is used. Count- near the blanks. ing-efficiency data and background data from 6.1.5 Allow prepared samples to remain standardis and blanks run with each set are in the dark in the liquid scintillation spec- also determined. The new data from each set trometer for several hours before counting are averaged into the data from the preced- begins. This allows decay of fluorescence and ing sets, thus creating a moving average chemiluminescence. A minimum of 8 hr value for counting efficiency and background. standing is required with Instagel and 12 hr As new efficiency and background data ap- with 3A70*. pear they displace older data entered into the 6.1.6 Count each vial five times for 100 moving average. Data from four or five valid min. Total counting time for one sample is runs (no quenching or other aberration) en- 500 min. Counting time is reduced for very ter into the moving average used at any given active samples. One million counts is cutoff time. setting. 7.1 Triitium efficiency factor (E). Use equation 2: 6.1.7 Program the instrument to per- - form the external-standard ratio analysis E= Cn on samples, standards, and blanks at the dn(e-%) ' (2) 66 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS where 7.3 Cmversion of tritium concentration i - G, = average count rate of standard in pCi/l to tritium units. Tritium mncentra- (cpm) corrected for background pCi tritiudl and blank, tion in Tu= 3.22 d,= disintegration rate of standard (dpm),- corrected for blank (S --my 8. Repart h='decay constant of tritium (4.685 Concentrations in both tritium units and x~O-~month-'), picocuries per liter are reported to two sig- t, = elapsed time between certification of nificant figures dolwnto the minimum detec- the standard and hime of coun4 in tion level (MDL). The latter can only be same units a+s A, estimated because of the very pronounced where- effect of altitude on the background count. S= moving average count rate At sea level the MDL is estimated at 60 Tu, - (cpm) ob the standards,and and at 5,000 ft (1,500m) it is estimated at B = moving average count rate 150 Tu fos the direct .count methold. (cpm) of the blanks. '7.2 Calculation of trttium concentration : 9. Precision Use equation 1. The chemical recovery factor Precision is dependent on altitude in the (f) is an enrichment factor when electro- same way as MDL.At 500 Tu reproducibility lytic enrichment is applied to the sample. is approximately f. 20 percent. Precision im- Since electrolysis is not used the value of f proves with increasing concentration. is unity.

1,000~ i pCi tritium/l= 7 (1) Reference KVEf(e-it) where f is electrolytic enrichment factor de- Schroder, L. J., 1971, Determination of tritium in termined by standard included in the run, water by the U.S. Geologicai Survey, Denver, Colo.: U.S. kl. Survey Rep&. USGS-474-134, cpm/ml after electrolysis 22 p.; Avail. only from U.S. Dept. Commerce, f= Natl. Tech. Inf. Service, Springfield, 22151. cpm/ml beforre eleckrolysis * VA Tritium Liquid scintillation method, Reston lab (R-1173-76)

Parameters and codes: Tritium, dissolved (pCi/l): 07005 Tritium, in water molecules (Tu): 07012

1. Application on to the energy analyzer and scalar cir- The technique can be used directly to an- cuitry. alyze waters containing more than 60 Tu When liquid scintillation counting is used (190 pCi/l) . Many natural-water samples to determine a radionuclide in aqueous solu- contain less than 60 Tu so the tritium in tio'n,the water sample is dissolved or dis- such samples is normally enriched by elec- persed in a larger volume of organic solvent trolysis (method R-1174-76) before they are containing the scintillating chemical. A analyzed by this technique. widely used mixture for aqueous solutions The direct liquid scintillation counting was dioxane-naphthalene containing the method, while useful for analysis of the tri- scintillator 2,5-diphenyloxazole (PPO) and tium introduced into water by rainout in a secondary scintillator used to shift the some samples, is primarily used for measur- wavelength of the scintillations to the most ing tritium introduced in tracer tests and sensitive spectral region of the PM tube. locally by nuclear power and waste disposal Much early tritium data was determined by facilities. use of the above scintillation mixture. The dioxane-PPO combination has been super- 2. Summary of method seded by proprietory scintillators that pro- Liquid scintillation counting is based on duce gels when mixed with water in proper the conversion of the energy of a particle ratios. The newer scintillators have approxi- emitted by a radioactive nucleus to light mately doubled the sensitivity of liquid scin- energy by means of a scintillating chemical. tillation counting for tritium. These solu- The scintillations are detected by a photo- tions generally include an organic solvent multiplier (PM)tube. The electrical signals (toluene or xylene, for example) in which from the PM tube are amplified and sent the scintillator is dissolved, plus a strong through a two or three channel analyzer liquid detergent to promote emulsification be- where sorting by energy ranges takes place. tween the sample water and the scintillator- The counts in each channel are displayed on containing organic solvent. a scaler and are printed and(or) read out Since the exact composition of the pro- on punched tape or magnetic tape. In order prietory mixtures is not available, these are to minimize spurious counts, from cosmic listed by trade name in the reagent section. radiation or electrical noise inherent in the While the mechanism of liquid scintilla- PM tubes themselves, for example, two PM tion counting is not completely understood, tubes are used to detect the scintillations. it seems clear that the energy transfer is The two tubes are operated in concidence, so a two-stage process, with initial energy that only if a scintillation event is detected transfer to the solvent followed by transfer by both tubes simultaneously is a signal sent from the solvent to the scintillator. Many

67 68 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS substances, including water, interfere in the results. Radon has a short half-life (3.8 d) energy transfer process to quench the scintil- mdits presence is obvious from a decline in lations and reduce the count rate. Excessive :ount rate while a sample is counted over a salt content, certain metals, and organic com- period of several days or a week. pounds quench in varying degree. Colored A radon-containing sample can be puri- substances may quench by light absorption fied by distilling it to remove any parent in addition to interfering with energy trans- radium-226, allowing the sample to stand fer. Quenching substances are generally re- for 34weeks for the radon to decay, and moved by vacuum distillation. redistilling before counting to remove the The quenching effect of water is compen- radon daughter lead-210. sated by using a constant volume of water The analysis of tritium in highly radio- and constant ratio of water sample to scin- active samples, such as those from nuclear tillator mixture in both samples and stand- facilities, may be confused by the presence ards. The count rate of a particular sample of other radioactive noble gases, particularly depends on both the volume of sample and -85 and argon-41. Rhinehammer and the ratio of liquid scintillator mixture to Lamberger (1973) discuss techniques for water sample. As the fraction of water sam- tritium analysis in the presence of concen- ple in the water scintillator mixture in- tration of other radioisotopes much higher creases, count rates increase until a point is than found in natural samples. reached where the quenching effect of the additional tritiated water exceeds the effect Samples prepared for counting by elec- of the increased radioactivity. It is found trolytic enrichment are normally free of radioactive or quenching interfering sub- that a practical compromise between maxi- mum sample volume and scintillator volume stances. provides optimum sensitivity. The ratio of sample to scintillator solution (in a constant 4. Apparatus total volume) can vary slightly without af- 4.1 Counting equipment, liquid scintilla- fecting the counting characteristics of the tion spectrometer, with two PM tubes, oper- mix. Thus optimally sensitive mixes can be ating in coincidence minimum of two chan- produced in routine work. nels for pulse-energy analysis ; automatic The permissible ratio of water sample to sample changer,minimum 100 samples ; con- liquid scintillator mix is also controlled by stant temperature chamber for PM tubes physical stability of the gel formed. A de- and shield and sample changer, adjustable to tailed description of the method is given by as low as 0°C; readout device(s) to paper Schroder (1971), and the counting charac- for visual inspection of results and to teristics and thermal stability of various punched tape or other automatic data proces- water-scintillator solution mixtures are sing (ADP)compatable form for data trans- given in the manufacturers’ literature. fer to computer for final calculations. Opltional : automatic external standardiza- 3. Interferences tion by channels-ratio method for quench de- Potential interferences come from other termination. Properly prepared samples do radionuclides or quenching subytances pres- not differ in quenching, and this determina- ent in the sample. Both types are removed tion is unnecessary. The presence of the ex- by vacuum distillation of the sample before ternal standard source near the counting preparation of the counting mixture. chamber may add additional background. A rare ground-water sample may contain Counters as received from the manufac- radon-222 in an amount large enough to per- turer will not normally be adjusted for sist in the sample even after distillation. optimum counting of water mixtures, and Some radon decay is counted in the tritium before routine tritium measurements are be- energy channel and to spuriously high must be carefully adjusted in the laboratory METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 69

gun. Adjustments are possible to: PM-tube standard tritiated water with “dead” water, high voltage ; amplifier gain ; energy analyz- that is, water containing less than 1 Tu. A ing channel limits; and (by a manufacturer’s standard containing 50 to 100 dpm/ml tri- representative) the time interval in which a tium (7 to 15X103 Tu) is adequate. pulse in one PM tube is taken as simultane- 5.3 Water, “dead”. Water with no meas- ous with a pulse in the other and sent on to urable tritium (“dead” water) is required the analyzer circuitry. for determination of counter-background The type of scintillator solution and opti- rate and for dilution of standard tritiated mum water to solution ratio of the counted water. It is very difficult to c6nfirm €hat a mixture must also be chosen, and the count- water is truly “dead,” and it is usually neces- ing-chamber temperature set as low as pos- sary to make the assumption that water from sible to minimize F’M-tube noise without a deep well in a confined aquifer a hundred causing the counted mixture to separate or kilometers or more from the recharge area freeze. is “dead.” This assumption can be correct The utility of a liquid scintillation spec- only if the well is pumped sufficiently to ex- trometer for low-level counting is deter- pel all meteoric water that may have entered mined by its background count rate and its the well and surrounding aquifer by leakage. tritium-counting efficiency. A frequently used counting mixture consists of 12 ml 6. Procedure H,O sample plus 13 ml InstarGel (Packard 6.1 Distillation. Samples enriched by Instrument Go.), in a polyethylene vial. electrolysis are distilled in the final step of Presently available (1976) commercial that procedure and require no further treat- counters are capable of counting this mix- ment before preparing the counting mixture. ture, with efficiences of 20 percent or greater, All other samples must be vacuum distilled and backgrounds of 3.6 cpm or less. Older before preparing the counting mixture. instruments, with efficiences as low as 12 6.2 Counting. Samples are counted in percent and backgrounds as high as 3.8 cpm, sets, each requiring about 1 week of count- may also be satisfactory, but will require ing. A set typically includes 1 standard, 1 longer counting times. or 2 blanks (backgrounds), and from 7 to 4.2 Counting vials, commercial, screw-cap 15 samples. The number of samples per set polyethlyene liquid scintillation counting depends on tritium content. Low-tritium vials, with caps with reflective liners. samples require longer counting times, and 4.3 Vacuum-distillation apparatus. so fewer can be counted per week. 4.4 Miscellaneous glassware, pipets for Members of a count set are prepared to- preparing counting mixtures, standard solu- gether and in the same way. The proper tions, and so forth. volumes of sample, standard, or “dead” 4.5 Analytical balance. water is pipetted into a tared counting vial and its mass determined to kO.01 g. The 5. Reagents scintillation solution is then added by manual 5.1 Scintillator, commercial solutions for or autopipet. The water to scintillator solu- counting water mixtures, such as :Insta-Gel, tion ratio must be constant within a count Packard Instrument Co.; Ready-Soln, Beck- set for the optimized counter settings. If suf- man Instrument Go. ; Scintillator 3A70*,Re- ficient sample volume is not available, fol- search Products International ; Scinti Verse, lowing electrolysis, for example, the mass of Fisher Scientific Go. ; Aquasol, Nuclear Asso- sample in the counting vial is first measured, ciates ; or ScintillAR, Malinkrodt Chemical then “dead” water added to bring the total co. water volume to that required by the counter 5.2 Tritium standard solution: Appro- settings. priate standards for determining counter ef- The vials are capped and the contents ficiency are prepared by the dilution of NBS thoroughly mixed. Heating the mixture to 70 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS - 4O0-6O0Cpromotes emulsification,but is not B = average count rate (cpm) of the ; necessary. The set is then placed in the blank (s) , (backgrounds) in counter sample changer for at least 24 hr the set. before counting begins to reach thermal 7.2 Calculation of tritium concentration. equilibrium and for chemiluminescence (if Use equation 1: - any) to decay. C Members of each set are counted sequen- TU=---- , (1) tially for a preset time (40 to 100 min) or KVEf (e-ht) until a preset number of counts (5,000 to where- 10,000) have accumulated. The counter then C=average sample count rate above records the sample number, count time, and background (cpm) , counts accumulated, and moves the next sam- V=mass of sample water counted (g) , ple into position. After the last sample, the E= counting efficiency, from equation 2, sample changer returns to the first sample and the cycle repeats. If the automatic external-standardization [El9 option on the counter is used, the standard- f = electrolysis enrichment factor. For ization is done at the end of each count on samples counted directly, f = 1 each sample and the results recorded before (for calculations of f for elec- changing to the next sample. trolyzed samples, see method R- The total time required for counting a 1174-76), sample is determined by its tritium content A = decay constant of tritium (1.534 (see see. 7.3 below). If a set contains several xlO-* d-l), high-tritium samples, these can be removed t= time elapsed between sample collec- individually when they have counted long tion date and date counted, in enough, and counting of the remainder of same units as A, and i the set continued. K=7.13X10-3 dpm/g Tu. 7.3 Calculation of tritium counting error. 7. Calculations The error term which accompanies tritium 7.1 Tritium counting efficiency (E).Use results is calculated such that there is a 67 equation 2: percent probability that the true tritium content of the sample is in the range of the E= e, reported value +- the error term. d, (e-%) X v,' Errors of tritium analyses are due pri- where marily to the uncertainty inherent in any - en= average count rate of standard attempt to measure the rate of occurence __ (count rate) of a random process (radio- (cpm) above background (5'-B), active decay). For a count rate, R, the d,= disintegration rate of standard standard deviation v=R/t,where t is the (dm/g), total count time. Errors in counting both V,=mass of standard counted (g) , the sample and background are included in X=decay constant of tritium (1.534 the n& sample count rate c, above. That X10-4 d-I) , is, if C=S-B, t, = elapsed time between certification of the standard and time of (3) count in same units as A, In sets counted as described above,the count- where- ing time for background and sample are the S=average count rate (cpm) of same, so equation 3 can be written: the standard in the set, and crc= (C+2B) /t. (4) METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 71

The relative error, or precision of the count 8. Report rate measurement is__ : Tritium errors are reported to two sig- ,c/c. nificant figures or to the nearest 0.1 Tu, whichever is larger. The tritium result it- Thus equation 4 shows_- that analytical pre- self is reported to the same number of sig- cision increases (&/C hcomes smaller) nificant figures. with longer count time, or, for a fixed count Tritium data are frequently required in time, increases with higher sample count pCi/l rather than in Tu. To convert, use the rates (a higher tritium-counting efficiency expression : or a higher sample tritium concentration) or with a lower background count rate. PCi tritium/l=3.2 Tu. In addition to the count-rate error, small errors are associated with each of the other 9. Precision terms in equation 1, and the reported uTu is The precision of tritium analyses varies calculated with the expression with sample tritium content and with labora- tory configuration and location. The one standard deviation (1~)error is reported with each result. The calculation of this (5) error is described above (see. 7.3). There are also errors associated with the K and (e-ht) terms in equation 1, primarily References due to uncertainties in the knowledge of the Schroder, L. J., 1971, Determination of tritium in tritium half-life, but they are negligible for water by the U.S. Geological Survey, Denver, most purposes. Colo.: U.S. Geol. Survey Rapt. USGS-474-134, Because Tu-error calculations are tedious 22 p.; Avail. only from US. Dept. Commerce, Natl. Tech. Inf. Service, Springfield, VA. 22151. and complicated, they are done by computer Rhinehammer, T. B., and Lamberger, P. H., 1973, from the counting data punched or other- tritium content technology, U.S.A.E.C. report wise directly recorded by the counter. WASH-1269, 292 pp. (Avail. from NTIS).

Tri ti um E I ec t r o I y ti c e n r i c h m e n t-I i q u i d sc i n ti I I a ti o n mme t h od , Denver lab (R-1172-76)

Parameters and codes: Tritium, dissolved, (pCi/l): 07005 Tritium, in water molecules (Tu): 07012

1. Application techniques were introduced by Kaufmann Gas counting preceded by electrolytic an- and Libby (1954). richment is the most sensitive analytical Electrolytic enrichment is carried out by method for tritium and is applicable down to adding sodium peroxide (forms sodium hy- concentrations in the range of 1 Tu. With droxide), followed by carefully controlled very careful control of ambient laboratory electrolysis in specially designed cells. The tritium levels, and a valid blank based on cells of Ostlund and Werner (1961) are used. “dead” water, it is possible to,apply the meth- The isotopic fractionation factors are im- od to waters as low as 0.2 Tu. Where gas- proved by operation at low temperature. counting equipment is not available or where Hence, the electrolysis is carried out while a lower detection limit of 25 Tu is satisbac- the cells are partially immersed in a cold bath maintained at a temperature just above ~ tory, electrolytic enrichment followed by liquid scintillation counting permits the anal- freezing. ysis at lower concentrations than by liquid The percentage of recovery of tritium in scintillation alone (method R-1171-76) . The the electrolysis is a complex function of tem- technique also may be applied to samples hav- perature, current density, and electrode sur- ing high concentrations of nonvolatile radio- face reactions which are not fully understood. activity contamination such as strontium-90, Practical systems have been developed which because the sample preparation steps elimi- achieve 70-80 percent recovery of tritium in nate solid materials. electrolysis from approximately 500 to 10 ml. Xore extensive electrolysis provides greater The technique is generally applicable to enrichment but lower percentage of recovery. determine tritium introduced into water by The reproducibility of recovery between elec- rainout and to measure natural levels of tri- tium in surface waters. The technique is not trolyses is approximately 3-4 percent under sufficientlysensitive for the determination of normal conditions. Liquid scintillation counting is based on very low natural tritium levels. the conversion of the energy of a particle 2. Summary of method emitted by a radioactive nucleus to light energy by means of a scintillating chemical. When it is necesary to determine tritium The scintillationls are detected by a photo- at a lower concentration, with improved ac- multiplier (PM)tube. The electrical signals curacy, than is available in the liquid scintil- from the PM tube are amplified and sent lation method (R-1171-76), an electrolytic through a simple multichannel analyzer enrichment step is introduced ahead of the (three or four channels at most) where sort- liquid scintillation counting. The electrolysis ing into energy takes place. The counts in 74 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS each channel are displayed on a scaler and ;he ratio of liquid scintillator mixture t0 1 may be read out on paper tape, punched tape, water sample. As the fraction of water sam- or magnetic tape. ple in the water-scintillator mixture in- Liquid scintillation counting is used pri- xeases, count rates increase until a point is marily for the counting of beta emitters al- reached where the quenching effect of the though it can also be utsed for alpha-emitting additional tritiated water exceeds the eff ect isotopes. of the increased radioactivity. It is found When liquid scintillation counting is used that a practical compromise between maxi- to determine a radionuclide in aqueous solu- mum sample volume and scintillator volume tion, the water sample is dissolved or dis- provides optimum sensitivity. The maximum persed in a larger volume of organic solvent on the curve of activity versus volume of tri- containing the scintillating chemical. A wide- tiated water (in a constant overall volume) ly used mixture for aqueous solutions was is a rounded plateau. Therefore, the propor- dioxane-naphthalene containing the scintil- tion of water sample to scintillating liquid lator 2,5-diphenyloxazole (PPO) and a sec- mixture is not critical and is easily repro- ondary scintillator used to shift the wave- ducible in routine work. length of the scintillations to the most sensi- The permissible ratio of water sample to tive spectral region of the PM tube. Much of liquid scintillator mix is also controlled by the tritium data reported in the literature physical stability of the gel formed. was determined by use of the above scintilla- A detailed description of the method is tion mixture. The dioxane-PPO combination given by Schroder (1971). has been superseded by proprietory scintil- lators that produce gels when mixed with 3. Interferences water in proper ratios. The newer scintil- There are no interferences in the analyti- lators have approximately doubled the sensi- cal method when electrolysis is included. Dis- tivity of liquid scintillation counting for tri- tillation is used to remove both quenching tium. Since the composition ’of the proprie- substances and radionuclides that could con- tory mixtures is not available, these are listed tribute to excess counts. Distillation and by trade name in the “Reagent” section. electrolysis are fully effective in removing While the mechanism of liquid )scintilla- inorganic salts, high-boiling organic com- tion counting is not compldely understood,it pounds, and gaseous radioisotopes. Krypton seems clear that the energy transfer is a two- and all other gases are Istripped out in the stage process, with initial energy transfer to electrolysis owing to prolonged bubbling of the solvent followed by transfer from the sol- and hydrogen through the sample. vent to the scintillator. Many substanceis, in- Further protection against interferences cluding water, interfere in the energy-trans- is provided by the energy discrimination in fer process to quench the scintillations and the liquid-scintillator analyzer and the ex- reduce the count rate. Excessive salt content, ternal standard-ratio test. certain metals, and organic compounds quench in varying degree. Colored substanca 4. Apparatus may quench by light absorption in addition to interfering with energy transfer. Quench- 4.1 Autopipet, 25-ml maximum capacity. ing substances are generally removed by 4.2 Counting equipment. Liquid scintilla- vacuum distillation. tion spectrometer ; counting systems capable The quenching effect of water is compen- of meeting the follo’wingspecifications : sated by using a constant volume of water Background-not to exceed 5 cpm at sea level and constant ratio of water sample to scin- in tritium channel. tillator mixture in both samples and stand- Counting efficiency-not less than 24 percent ards. The count rate of a particular sample with optimum sample-scintillator mixture and depends on both the volume of sample and polyethylene vials. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 75

Sample capacity-at least 100 samples. 4.4 Pipets, 8-ml. Operation-programmable and automatic. 4.5 Vacwum-distillation apparatus. Con- Internal check-an external standard and ratio computation capability required. sists of a 100-ml round-bottom flask as the Readout-automatic printout. distillation flask and a 125-ml round-bottom 4.3 Electrolysis unit. Contains the fo1lo;w- flask as the condenser flask. The distillation ing components: flask is heated with a rheostabcontrolled 4.3.1 Electrolysis cells. See figure 8. The mantle and the condenser flask dips into a Ostlund cell has a mild-steel cathode (where Dewar containing isopropanol-dry ice. The reduction of hydrogen isotopes occurs) and two flasks are connected by a 20-mm diame- stainless-steel anode. The glass envelope is ter U-tube,10-ern long with ground-glass con- designed to attach directly to a vacuum-dis- nectors and stopcock for application of tillation apparatus. vacuum. Heating tape is coiled around the 4.3.2 Power supply, direct current, at U-tube connecting the two flalsks. least 6 amperes at 30 volts. 4.3.3 Freezer, floor-model,large enough 5. Reagents to hold two rows of five electrolysis cells. 5.1 Scintillator, Instagel (Packard Instru- 4.3.4 Exhaust lines, to vent the explosive ment Co.) for low-te8mperature counting mixture of oxygen and hydrogen generated (1”-4”C). A preblended gel-forming scintil- in electrolysis. lator designated 3A70” (Research Prod- ucts International) for room-temperature counting. 5.2 Sodium peroxide, reagent-grade. 5.3 Tritium standard solution: Appropri- ate standards are prepared by the dilution of NBS standard tritiated water with “dead” water, that is, water containing less than 1 Tu. 5.4 Water, “dead”: The tritium blank in- troduced by reagents and leakage during electrolysis must be tested at intervals by analyzing “dead” water (water with no measurable tritium content) , in exactly the same procedure as for a normal water sam- 500rnl SAMPLE RESERVOIR ple. It is very difficult to confirm that a water lOOrnl CELL BODY is truly “dead,” and it is usually necessary to ELECTRODES SEPARATED make the assumption that water from a deep BY TEFLON SPACERS well in a confined aquifer a hundred kilome- D - LEAD-IN WIRES ters or more from the recharge area is E - TEFLON SPACERS “dead.” This assumption can be correct only F - CONNECTOR FOR GAS if the well is pumped sufficiently to expel all EXHAUST TUBE AND meteoric water that may have entered the ELECTRICAL ENTRY well and surrounding aquifer by leakage.

6. Procedure 00 6.1 Electrolysis. (In electrolysis and all other steps, heat-dried glassware is used.) 6.1.1 Distill 55-mlvolume of water sam- ,I ple in the vacuum-distillation apparatus. The Figure 8.-Ostlund electrolysis cell. unit is evacuated before use, and the water 76 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS sample is protected from atmospheric mois- tion apparatus and apply vacuum. The ture during distillation. Distill to dryness. sample first bubbles to release trapped gases Recovery is usually slightly less than the and then freezes. Apply a heat lamp or heat original volume because of water of hydra- gun to distill the sample into the liquid- tion remaining in the salts residue and drop- nitrogen-cooled receiving bulb. Weigh the lets on the walls of the apparatus. The triti- bulb after completion of distillation to deter- um fractionation attributable to nondistilled mine the volume of the water sample water is insignificant. collected. 6.1.2 Transfer the distillate to a clean 6.2 Counting. and dry Bstlund electrolysis cell (fig. 8), add 6.2.1 Pipet 8.00 ml of the distillate from 0.75 g of sodium peroxide, add 50 ml of the the preceding step into a 25-ml polyethylene distilled-water sample, and stopper the cell. vial, and add 14 ml of scintillator mix. The An argon atmosphere is maintained in the choice of scintillator depends on the type of reservoir to eliminate contact with atmos- liquid scintillation spectrometer to be used pheric moisture. for counting. Instagel is used with the instru- 6.1.3 Prepare a blank sample (“dead” ments that aperate at cold temperature water) for electrolysis using the above pro- (3°C) and 3A70* is used with the instru- cedure. Prepare a standard for electrolysis ments that operate at room temperature. Cap (200 Tu is a convenient concentration range) the vial and mix. Heat the Instagel-sample using the above procedure. mixture on a hotplate at approximately 6.1.4 Set up one blank, one standard, 100°C for 2 o’r3 min. This clarifies the mix- and four samples in series in the electrolysis ture. The 3A704-sample mixture does not bath. A larger number of samples may be require heating. The above operation is car- used if sufficient voltage is available from ried out under subdued red light to filter out the electrolysis power tsupply.The tritium the blue region of the spectrum. This mini- enrichment for one group of samples is deter- mizes excitation of fluorescence background mined by the enrichment of the standard in in the sample. series with the group. 6.2.2 Prepare three blanks and two 6.1.5 Proceed with the electrolysis after standards in the same manner as the Sam- samples have been cooled in the cold chest. ples. Place one standard in the 2d counting The temperature of the ethylene glycol-water position in the spectrometer and one in the coolant bath is maintained at 0°C throughout 10th po’sition. Place blanks at intervals the process. Electrolysis from 50 ml to ap- throughout the run of 10-14 samples. proximately 10 ml using 4-ampere current. 6.2.3 Place three sealed standards (tri- 6.1.6 Neutralize the highly caustic solu- ated toluene in glasssealed scintillator solu- tion in the cells to permit full recovery of tri- tion) in the group. One standard goes in the tium and to avoid mechanical and corrosion first counting position t~ permit monitoring problems in subsequent steps. Disconnect a the instrument before counting the samples. cell from the electrolysis line, and while pro- The remaining sealed standards are placed tecting from the atmosphere insert a dis- near the blanks. posable pipet through the hole in the Teflon 6.2.4 Allow prepared samples to remain spacer. Bubble carbon dioxide through slow- in the dark in the liquid scintillation spec- ly. About 5 min and 1 liter of carbn dioxide trometer for several hours before counting are required to complete neutralization. A begins. This allows decay of fluorescence and precipitate of sodium carbonate forms, chemiluminescence. A minimum of 8-hr 6.1.7 Distill the neutralized sample into standing is required with Instagel and 12 a small tared Pyrex bulb using vacuum-dis- hours with 3A70”. tillation apparatus. The sample is cooled (but 6.2.5 Count each vial five times for 100 not frozen) in liquid nitrogen. Attach the min. Total counting time for one sample is still-liquid sample to the inlet of the distilla- 500 min. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 77

I 6.2.7 Program the instrument to per- form the external-standard ratio analysis on samples, standards, and blanks at the end of where the counting run (see instrument instruc- - tion manual). This procedure is a check for c, = average count rate of standard quenching. The ratio of low-energy counts in (cpm) corrected for background channel A of the spectrometer is established and blank, for both tritium and an external standard d, = disintegration rate of tstandard placed near the sample vial. The value of R (dpm), corrected for blank (IF- in the following equation should be constant -B), to within 50.2. h=decay constant of tritium (4.685 x month-I), R=[ Counts in A tritium] t, = elapsed time between certification of Counts in B the standard and time of count in standard]. same units as A, where- Individual samples that fall ontside this S = moving average count rate range must be reanalyzed. If all values of R - (cpm) of the standards, and fall outside this range, the scintillator has B=moving average count rate deteriorated and must be replaced. (cpm) of the blanks. 7.2 Calculation of tritium concentration : 7. Calculations Use equation 1. The chemical recovery factor Several statistical schemes have been pre- (f) is an enrichment factor when electrolytic sented for the calculation and verification of enrichment is applied to the sample. tritium data, each intended to optimize a par- I ticular analytical situation. In the present analytical procedure the repeated counting of individual samples has the effect of aver- where f is electrolytic enrichment factor de- aging out short-term fluctuations. Statistical termined by standard included in the run, checks have shown that highest precision, in cpm/ml after electrolysis

this analytical situation, is attained by the f= cpm/ml before eleckdysis' use of longer term average values for the 7.3 Conversion of tritium concentration samples and standards. in pCi/l to tritium units. Counting efficiency and background values pCi tritium/l are determined with tritium standards when- Tritium concentration in Tu= ever a new lot of scintillator is used. Count- 3.22 ing-efficiency data and background data from 8. Report standards and blanks run with each set are Concentrations in both tritium units and also determined. The new data from each set picocuries per liter are reported to two sig- are averaged into the data from the preced- nificant figures down to the minimum de& ing sets, thus creating a moving average tion Ievel (MDL). The latter can only be value for counting efficiency and background. estimated because of the very pronounced As new efficiency and background data ap- effect of altitude on the background count. pear they displace older data entered into the At 5,000 ft (1,500m) it is estimated at 25 Tu moving average. Data from four or five valid for the liquid scintillation counting of an runs (no quenching or other aberration) en- elec.trolytically enriched sample. ter into the moving average used at any given time. 9. Precision 7.1 Tritium efficiency factor (E). Use Precision is dependent on altitude in the equation 2: same way as MDL.At 500 Tu reproducibility 78 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS is approximately 220 percent. Precision im- Ostlund, H. G., and Werner, 0. E., 1962, The el=- proves with increasing concentration and is trolytic dchment of tritium and deuterium for improved by electrolytic concentration of natural tritium measurements, in Tritium in the samples with lower tritium concentration. physical and biological sciences, Symposium Reproducibility for samples enriched by Proceedings (STI/PUB/39) IAEA, Vienna, p. electrolysis is limited by the reproducibility 95-104. of electrolysis (approximately -+ 3 pcmxnt) . Schroder, L. J., 1971, Determination of tnitium in References water by the U.S. Geological Survey, Denver, Kaufmann, S., and Libby, W. F., 1954, The natural (3010.: U.S. Geol. Survey Rept. USGS-474-134, distribution of tritium, Physical Review, v. 93, 22 p.; Avail. onZy from U.S. Dept. Commerce, p. 1337-1344. Natl. Tech. Inf. Service, Springfield, VA 22151.

I Tritium Electrolytic en richment-li quid scintilla tion method, Reston lab (R-1174-76)

Parameters and codes: Tritium, dissolved (pCi/l): 07005 Tritium, in water molecules (Tu): 07012

1. Application and soft iron electrodes. (See fig. 8, The limit of detection of tritium by the method R-1172-76, for diagram of electroly- liquid scintillation counting method (R- sis cell.) During operation, the cells are kept 1173-76) is about 60 Tu (190 pCi/l) .Many at O0-loC to improve electrolytic tritium re- surface-water and moat ground-water sam- covery and to, minimize loes of sample by ples contain less than 60 Tu, and cannot be evaporation. A maximum of 100 ml is elec- analyzed directly. Concentration of the triti- trolyzed in the cell. If larger samples are to um in low-tritium samples by electrolysis be- be electrolyzed, the sample is periodically fore counting by liquid scintillation permits added to the cell from the reservoir as elec- the analysis of waters containing as little as trolysis proceeds. 1 Tu (3 pCi/l) . Electrolysis proceeds at 6 amperes until less than 25 ml remains. Then the current is reduced in steps to as low as one-half ampere 2. Summary of method as the remaining volume decreases to the de- When water is decomposed to H, and 0, sired 5-ml final volume. Electrolysis from a gas by electrolysis there is a strong isotope starting volume of 100 ml requires about 4 d. fractionation effect which results in the Following electrolysis, the sample water is heavier isotopes (tritium in particular) be- separated from the electrolyte by vacuum ing concentrated in the remaining liquid distillation and is then ready to be counted. phase. Under the proper conditions, recovery Electrolysis is performed in sets of cells of more than 70 percent of the initial tritium connected in series to a constant voltage, cur- is possible. Thus, if a sample is reduced from rent-limiting power supply. Each set includes 500 ml to 5 ml by electrolysis, the tritium in an electrolysis standard, a blank, and from the residual 5 ml will have been concentrated 4 to 10 samples. 500 x 03 by at least: =70 times. When such 3. interferences 5 Samples are distilled before and after elec- an electrolyzed sample is counted using the trolysis, and are thoroughly gas-stripped by procedure described in method R-1173-76, the H, and 0,produced during the elwtroly- tritium levels as low as 1 Tu can be detected. sis itself. Thus all potential interfering sub- The electrolysis procedure used is essen- stances are effectively removed,and no radio- tially that described by Ustlund and Werner active or other interferences remain in the (1962). The sample, after distillation, is sample ready for counting. made basic with NaOH or Na,O, and elec- The major interference with low-level tri- trolyzed in glass cells (Ustlund cells) with tium analysis is contamination by tritium it-

79 80 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS self. The electrolyte may contain tritium, or nethod R-1173-76) is calcuiated using equ+ the sample may pick up tritium during exces- ;ion 1: sive exposure to the laboratory atmosphere. To minimize contamination, it is important that sources of tritium above levels naturally present in the air be rigorously excluded vhme from the laboratory. These sources include Vo= volume of sample before electrolysis, luminous watches and such high-tritium sam- Vf= volume of sample following electrol- ples as may result from tracer tests. ysis, and Samples of “dead” water should be run p= the separation factor. through the entire electrolysis and counting The electrolysis standard with each set is process as blanks to monitor contamination :ounted as if it were a sample and its tritium and to permit corrections to be made to the content determined. Then : final reported tritium content (see section TUf below). fNtd=Tx

4. Apparatus where The apparatus required is the same as that Tuf=Tritium content of electrolysis described in method R-1173-76 with the ad- standard after electrolysis, dition of the following : Tu,= Tritium content of electrolysis ostlund electrolgsis cells. (See fig. 8,meth- standard before electrolysis, and od R-1172-76.) fstd= standard enrichment factor. Vacuum-distillation apparatus, (1) for The standard enrichment factor is then predistillation, capable of handling volumes substituted for “f” in equation 1 and p is up to 500 ml, and (2) for distillation after calculated for the set. \ electrolysis, capable of handling from 3 to 10 ml. 7.2 Calculation of tritium concentration. Calculate tritium concentration in the sample Freezer for cooling electrolysis cells. in the manner described in method R-1173- 76. 5. Reagents As specified in method R-1173-76 with 8. Report the addition of the following : Report as described in method R-1173-76. Electrolyte, NaOH or Na,O,. Electrolysis standard, a tritium standard 9. Precision solution prepared as described in method R-1173-76, sec. 5.2, and containing about 1 There are uncertainties in each of the dpm/ml. terms in equation 1 leading to f. These are the weighing errors in V, and V,, and varia- tions in the electrolysis proms itself giving 6. Procedure rise to variations in p. Experience suggests Perform electrolytic enrichment as de- that the one standard deviation error of f scribed in sec. 2 above, then follow procedure (ofif in equation 5, method R-1173-76) is described in method R-1173-76. about 5 percent. There is also error associated with the 7. Calculations value of the blank which is used to correct for 7.1 Enrichment factor. The enrichment sample contamination by reagents and ex- factor, f, required to calculate the sample tri- posure in the laboratory. Under favorable tium content from the count data (see conditions, the error in the blank may be as METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 81

1974). In routine work, the error in the Gstlund, H. G., and Werner, E., 1962, The dectrolytic blank is taken as equal to the value of the enrichment of tritium and deuterium for natural blank correction itself-usually from 0.1 to tritium~. measurements: in Tritium in the Physi- 0.3 Tu. Thus, although it is sometimes possi- cal and Bidogical Sciences; Internat. Atomic ble to count a sample to a precision of less Energy Agency, Vienna, V.I., p. 95-104. than 0.1 TU, the real precision of routine trim GsuUd, H*G-7 DomeY, H*G.7 and Rooth, C-G.9 1974, Geosecs North Atlantic radiocaTbon and tritium tium analyses is limited by the blank error to results: Earth and Planet. Science Letters, v. 20.1 to -1-0.3Tu. 23, p. 69-86.

U raniu in, dissolved Fluorometric met hod-d i rec t (R-1180-76)

Parameters and codes: Uranium, dissolved (pg/l): 22703 Uranium, dissolved (pCi/l): 80010

1. Application centration, quench-compensation techniques may be used. When quenching elements are The method is suitable for the determina- tion of uranium in nonsaline waters present in relatively high concentration, it is necessary to purify the uranium by extrac- (<10,000 mg/l dissolved solids) in which tion. This technique described as “Fluoro- uranium fluorescence is quenched less than 30 is metric method-extraction procedure” (R- 30 percent. If quenching exceeds percent, it 1181-76). is advisable to use the extraction method. The latter method is much more time con- Although the fluorescence of the sample is suming. Therefore, it is usual practice to directly proportional to the uranium concen- apply the direct fluorimetric method as a first tration (disregarding quenching effects), it step unless previous analysis of samples from is not possible to use a constant calibration a particular area has shown that the simple straight-lineplot of fluorescence against con- approach is not possible. centration. This is a result of variation in I The minimum detection limit varies with properties between batches of flux, variations the properties of the sample, flux, and fluori- in the fluxing temperature, possible surface meter, but is normally 0.3 pg/l. oxidation during fluxing, and variations of uranium impurity in different batches of flux. 2. Summary of method The above effects are minimized by running uranium standards with each set of samplels The fluorimetric method of determining so that a new calibration of concentration uranium is among the most sensitive and spe- against fluorescence is made under the condi- cific of analytical methods. The intense fluor- tions existing for each set of analyses. escence of uranium when fused in a sodium fluoride-sodium carbonate-potassium carbon- The materials used for the preparation of ate flux is utilized to determine quantitative- the flux always contain a small amount of ly the amount of uranium present in the uranium. Fluorescence from this source plus sample. In its simplest form the analysis is reflected light not absorbed by the filters in carried out by fusing a dry residue of the the fluorimeters make up the blank. The evaporated-water sample in fluoride-carbon- fluorescence component of the blank is sub- ate flux, allowing this to solidify into a small ject to quenching while the reflectance com- disk, and determining the fluorescence under ponent is not. The blank for a highly ultraviolet light in a reflection-typefluorime- quenched sample is, therefore, less than the ter. , copper, manganese, and a blank for a sample with relatively low few other elements in water quench the fluor- quenching. A graphical method of compen- escence in varying degree. When the sating for this effect on the blank value was quenching elements are in relatively low con- developed by Thatcher and Barker (1957).

83 84 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS The following method is similar to that described in Water-Supply Paper 1696-C (Barker and others, 1965) for the determina- tion of uranium in nonsaline water.

3. Interferences Direct spectral interference is not a prob- lem in the fluorimetric method. Cadmium fluoresces in high carbonate flux disks at ap- proximately the same wavelength as uranium (Booman and Rein, 1962, p. 102), but inter- ference from this source is unlikely in most natural waters. High concentrations of salt cause difficulty in the preparation of the flux disks. The quenching effect of transition metals has been cited above.

4. Apparatus 4.1 Crucible tongs, platinum tipped for holding hot platinum dishes. 4.2 Fluorimeter. A reflection-type instru- ment of high sensitivity equipped with a D sample carriage to accept small disk-shaped solid samples is required. The sample cavity should be approximately 35 mm in diameter and 5-mm deep. Instructions herein pertain A - ROTATING SAMPLE TABLE to the Jarrell-Ash Model 26-000 instrument, but any fluorimeter that fulfills the above re- quirements may be used. B - QUARTZ RODS TO HOLD 4.3 Fusion apparatus. The rotary fusion PLATINUM DISHES machine developed by Stevens and others (1959) and modified by Barker and others (1965) is used (fig. 9). A rotating sample C - RING BURNER carriage is mounted above a ring of burners and is slowly revolved during the fusion to assure that each sample receives the same D - SHEET METAL HOUSING heating. The samples are contained in plati- num fusion dishes resting on quartz rods. The fusion unit is tilted approximately 30" E - GEARED MOTOR during part of the fusion so that molten flux washes the sides of the fusion dishes to sweep down any sample residue that may adhere. F - LEVER FOR INCLINING The design of the burner must be adapted to BURNER 20' local gas composition and pressure to obtain Figure %-Stevens apparatus for fusion and mixing the optimum temperature. of sample and flux in uranium determination. 4.4 Fusion dishes. The dishes are fabri- cated of platinum to a shallow-dish shape ;y) compatible with adequate thickness for that provides maximum exposure of sample strength (fig. 10). The rounding of the bot- disk surface (for high fluorescence sensitivi- ;om permits easy removal of the solidified METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 85

5.4 Sodium fluoride solution, 1 ml=0.01 g NaF : Dissolve 10 g of dry sodium fluoride in distilled water and dilute to 1,000ml. 5.5 Uranium standard solution I, 1 ml =lo0 pg U: Dissolve 0.1773 g of reagent- grade uranyl acetate dihydrate in approxi- mately 500 ml of distilled water. Add 10 ml of concentrated nitric acid and dilute to 1,000 c ml in a volumetric flask. Store in a Teflon bottle. 5.6 Uranium standard solution 11, 1 ml =LOO pg U: Dilute 10.0 ml of uranium standard solution I to 1,000 ml with distilled water. Store in a Teflon bottle. DEPTH AT RIM - 10mm 5.7 Flux: Using anhydrous powdered re- DEPTH AT CENTER - 12mm agent-grade chemicals, weigh out 910 g of Figure 10.-Platinum dish for use in Stevens ap- Na,CO,, 910 g of K,CO,, and 180 g of NaF, paratus. and rough-mix in the glass-jar mill using a large porcelain spatula, or Lucite rod. Add flux disk. An identifying number is stamped the small Lucite rods, stopper tightly with a into the lip of each dish. polyethylene stopper, place the jar on the 4.5 Infrared drying lamps: Dual 250-watt mechanical rollers, and dry-mix overnight. infrared drying lamps in a pro+mtive metal shield are mounted on a ringstand for vari- 6. Procedure able heat adjustment. 6.1 Determination of flux constants :It is I 4.6 Micropipet, 50-pl capacity, Eppendorf necesary to determine r (the fraction of re- type,for addition of uranium standard. flected light in the blank) and f (the fraction 4.7 Mill. A 5- to 6-liter Pyrex glass-jar of fluorescent light in the blank) for each mill containing 15 em by 2.5 em cylindrical batch of flux. These “flux constants” are used Lucite rods is used to mix the flux. for all analyses made with the given batch of flux. Two calibration curves are prepared as 4.8 Pipets, volumetric 1-, 2-,3-, 4-, 5-, 6- shown in figure 11. The calibration curve X and 10-ml. is prepared with pure uraaium solutions and 4.9 Polyethylene jars, wide-mouth,4-liter the calibration curve Y is prepared with screw-capped,for storing flux. uranium solutions containing a constant amount of quenching agent. The two curves 5. Reagents intersect (P) at a negative uranium concen- 5.1 Chromium solution, 1 ml=30 pg Cr: tration. The interneetion of the curve X with Dissolve 0.085 g of K,Cr,O, in distilled water the fluorescence axis is point A, the un- and dilute to 1,000 ml. Weight and volume quenched blank. The intersection P of X and measurements need not be exact. Y is po,intB, the reflected light. The fraction 5.2 Copper solution, 1 ml= 60 pg Cu :Dis- of reflected light r is B/A and the fraction solve 0.236 g of CuSo4.5H,0 in distilled of fluorescent light f in the blank reading is water and dilute to 1,000 ml. Weight and l-r. volume measurements need not be exact. The procedure is as follows : 5.3 Manganese solution, 1 ml=20 pg Mn: 6.1.1 Measure 1 ml of sodium fluoride Dissolve 0.081 g of MnSo4.5H,0 in distilled solution and the following volumes bf urani- water and dilute to 1,000 ml. Weight and um standard solution I1 (microburet) into volume measurements need not be exact. platinum fusion dishes. 86 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

Solution II Dishes (ml) turn the table to the level position, and con- 1 and 2 ______0.000 tinue heating the melt for an additional 3 3 and 4 ______.020 min at the same temperature. 5 and 6 ______.040 6.1.6 Turn the heat cmtrol to the inter- 7 and 8 ______.060 9 and 10 ,080 mediate setting and heat at this tempera- ______ture for 3 min. 6.1.7 Turn to the low setting and heat for 3 min. Turn off the burner and allow the dishes to cool for 8 min with the fusion table still rotating. Finally, place the dishes in a desiccator, and cool for at least 30 min before measuring fluorescence. 6.1.8 Measurement of fluorescence: The following instructions apply to the Jarrel- Ash fluorimeter. Modify if other fluorimeters are used. Allow the instrument to warm up 30 min before use. 6.1.9 Set the fluorimeter reading to zero using the zero-adjustment knob. Push the i I I I I I I empty sample tray all the way in, depress .02 .04 .06 .08 .IO the X.01 key, and adjust to zero using the URANIUM CONCENTRATION (1-19) screw adjustment. Remove the blank flux X-FLUORESCENCE (UNQUENCHED) disk from the platinum dish by inverting it Y-FLUORESCENCE (QUENCHED) on a clean piece of paper. Place the disk in A-FLUORESCENCE BLANK (UNQUENCHED the fluorimeter tray and push into measure- PLUS REFLECTED LIGHT) ment position. Depress the X.l key and ad- B-REFLECTED LIGHT P-INTERSECTION POINT just the sensitivity so that a reading of 10 (on a scale of 100) is obtained for the blank. Figure 11.-Uranium calibrat‘ion curve. 6.1.10 Remove the blank disk and re- 6.1.2 To the even-numbered dishes, add check the zero setting. Use a soft-bristle 1 ml of chromium solution. brush to remove any particles sloughed into 6.1.3 Evaporate the solutions to dryness the sample tray from the preceding disk. If under the infrared lamps. Do not permit the the zero needs readjustment, repeat step samples to bake as this can result in loss of 6.1.9 until the empty holder reading is zero uranium, although the sodium fluoride ad- when the blank reading is 10. The “fluores- ded as the first step minimizes loss by over- cence” reading of the empty holder is minim- heating. ized by painting with a colloidal graphite 6.1.4 Fusion procedure: To each of the mixture such as Aqua-dag. This must dry dishes add 2 g of the flux mixture. Spread before use. Repainting is required at inter- and bank-the flux with a glass rod so that vals. any solids on the vertical walls of the dishes 6.1.11 Read the fluorescence of the will be covered. standard and sample disks using the X.1 6.1.5 Place the dishes on the rotating scale, if possible, or X1 and XI0 scales if table of the Stevens fusion apparatus, and needed. incline it by operating the positioning lever. 6.1.12 Plot the fluorescence of the disks Ignite the burner ring, adjust to maximum as a function of the weight of uranium (fig. heat, and heat the dishes until the flux is 11). Draw the best straight lines X and Y completely melted. This requires about 5 min. through the sets of points for the quenched Allow the fusion table to rotate in the in- and unquenched disks. Determine B, A, and clined position for an additional 2 min. Re- r and f as above. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 87

6.1.13 Repeat the calibration substitut- 7.2 Calculation of the corrected blank ing 1 ml of manganese solution for the ralue for each sample: The blank value to chromium solution. )e used with an individual water sample 6.1.14 Repeat the calibration substitut- nust be somewhat less than the blank meas- ing 1 ml of copper solution for the chromium ired instrumentally on the pure sodium solution. luoride-sodium carbonate disk because of 6.1.15 Average the values of r and f penching in the water sample of that por- determined for the three quenching elements ;ion of the blank value contributed by uran- as above. These mean values of r and f are urn impurity in the flux. Since a blank that used for all analyses using this batch of flux. represents conditions in the sample cannot 6.2 Analysis of the water sample. De prepared, it is necessary to calculate the 6.2.1 Four samples may be analyzed dank value for each sample on the basis simultaneously using the Stevens fusion ap- 3f the blank measured for the disk of pure paratus with 20 positions for samples, when fluoride-carbonate flux. The contribution of analyses are run in duplicate. One milliliter reflected light to the blank is assumed to be of sodium fluoride solution is added to the the same for the sample disks and the pure 20 platinum dishes. Two dishes serve as flux disk. The calculation is simply: blanks, Standard uranium (0.06 pg) is added Bs= (QfBr)+ (TBj) to each of two dishes, Four 7-ml aliquots of each sample are pipetted into four dishes. where Standard uranium (0.06 pg) is added to two B,= sample blank. of them. Bf=blank obtained with disk of pure 6.2.2 Proceed with the analysis as in flux, steps 6.1.3 through 6.1.11 above. f= fraction of fluorescent light for the batch of flux used, 7. Calculations r = fraction of reflected light for the 7.1 Determine the quenching factor, Q, batch of flux used, and for each sample from the equation: Q is defined in section 7.1. &=---,D-A - AI, 7.3 Concentration of uranium in the C-B AI, sample : This is calculated using equation : where Q = ratio of uranium fluorescence under quenching conditions to the fluorescence under no quench, where for a given sample, S =micrograms of uranium in the A = mean fluorimeter reading of un- standard, and spiked samples, V=sample volume in ml and the other B = mean fluorimeter reading of blank terms are as defined in sections disks, 7.1 and 7.2. C = mean fluoirimeter reading of stand- If it is known from previous experience ard disks, with water from a particular source that D =mean fluorimeter reading of sam- ple disks containing uranium the quenching factor Q is always greater than 0.7 (quenching less than 30 percent), spike, AI,= fluorescence increment of spiked it becomes possible to omit the determina- tion of and combine equations under 7.1 sample, and Q and 7.3 to obtain the following simplified = fluorescence of the standard. AI, expression : Note that the equation applies only when the amount of uranium spike in the sample pg/l of u= is equal to uranium in the standard. 88 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

8. Report alyzed with a relatively pure flux, and as Report concentrations less than 1.0 pg/l high as 0.5 pg/l when a highly quenched to one significant figure. Above 1.0 pg/l, re- sample is analyzed with a flux having rela- port two significant figures. Occasionally the tively high uranium impurity. Under aver- radioactivity is reported in pCi/l. The con- age conditions the MDL is 0.3 pg/l. version factor is 1 pg=O.33 pCi when only the radioactivity of uranium-238 is con- References sidered. Natural uranium contains 0.72 per- Barker, F. B., Johnson, J. O., Edwards, K. W.,and cent of uranium-235,and normally uranium- Robinson, B. P., 1965, Determination of urdum 238 is considered to be in equilibrium with in natural waters: U.S. Geol. Survey Water- an equal activity of uranium-234. The con- Supply Paper 16964, 25 p. version factor is 1 pg=O.68 pCi when all Booman, G. L., and Rein, J. E., 1962, Uranium in Kollthoff, I. M., and Elving, P. J., eds., Treatise three isotopes are included. on analytical chemistry, pt. 2 of Analytical chemistry of the elements, wc. A. Systematic 9. Precision analytical chemistry of the elements: New York, Intersci. Publishers, v. 9, p. 1-188. Precision is approximately rt MDL or k 15 Stevens, R. E., Wood, W. H., Goetz, K. G., and percent, whichever is larger. The MDL is a Horr, C. A., 1959, Machine for preparing phos- function of quenching, fluorescence intensity phors for the fluorimetric determination of uran- ium: Anal. Chemistry, v. 31, p. 962-964. of the flux, and uranium impurity in the Thatcher, L. L., and Barker, F. B., 1957, D~etermina- flux. It may be as low as 0.1 pg/l with a tion of uranium in natural waters: Ad.Chemis- sample having no significant quenching an- try, v. 29, p. 1575-1578. U raniu m, dissolved Fluorometric method-extraction procedure (R-1181-76)

Parameters and codes: Uranium, dissolved (pg/l): 80020 Uranium, dissolved (pCi/l): 8001 5

1. Application 3. Interferences The method is applied to water samples All interferences in natural waters are where the reduction of uranium fluorescence removed by the separation steps and have no by quenching exceeds 30 percent (as deter- effect. The method has not, however, been mined in “Fluorometric method-direct” extensively tested with industrial wastes, (R-1180-76) ), the concentration of total mine waters, and other waters that may have dissolved solids exceeds approximately 10,000 unusually high concentrations of heavy mg/l, or a minimum detection level lower metals. When such waters are encountered it than 0.3 pg/l is desired. is advisable to run a spiked sample contain- ing a known increment of uranium through 2. Summary of method the analytical procedure to test for possible Uranium is separated from quenching ele- residual quenching. A quenching correction ments and excesxive salt concentrations in a can then be made as in “Fluorometric meth- two-step separation procedure developed by od-direct” (R-1180-76) based on the per- Smith and Grimaldi (1954). Uranium is co- centage of reduction of the expected fluores- precipitated, as uranyl phosphate, on alumi- cence from the known increment of uranium. num phosphate from a large-volume water sample. Several quenching elements are car- 4. Apparatus ried down by the precipitate. Final purifica- 4.1 Centrifuge. tion is made by dissolving the phosphate pre- cipitate in dilute nitric acid and extracting 4.2 Centrifuge tubes, 40- or 50-mlcapac- with ethyl acetate or ethyl ether in the pres- ity Pyrex tubes with polyethylene screw-type ence of a salting agent. The organic solution caps. is evaporated to dryness in a platinum dish, 4.3 Evaporating dishes, Teflon, 125 ml. and the fluorescence is determined after fu- 4.4 Fluorimeter: See method R-1180-76. sion of the dry residue in sodium fluoride- 4.5 Fusion apparatus: See method R- sodium carbonate flux. 11 80-7 6. Barker and others (1965) modified the 4.6 Fusion dishes: See method R-1180- procedure slightly and evaluated the applica- 76. tion to natural waters. The present procedure 4.7 Micropipet, 50-pl-dispensing pipet, uses magnesium nitrate (Hellman and Wolf, Eppendorf type. 1952) to “salt-out” uranium from the nitric acid solution of the precipitate into the ethyl 4.8 Pipet atzd control, 5 ml. ether phase. The magnesium salt is slightly 4.9 Polyethylene jars: See method R- more effective than the aluminum salt and 1180-76. usually contains less uranium impurity,thus 4.10 Ultrasonic generator, 500-watt giving a lower blank correction. model.

89 90 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS

5. Reagents through steps 6.11 through 6.19 of this 5.1 Aluminum nitrate solution, 0.2 M: procedrrre. Dissolve 15 g A1 (NO,) -9H,Oin distilled wa- 6.2 Add 3 ml of concentrated HNO,,1 ter and dilute to 200 ml. ml of 0.2 M aluminum nitrate reagent, and 5 5.2 Ammonium hydroxide, concentrated. ml of 0.090 M diammonium-hydrogen phos- 5.3 Ammonium nitrate solution, 1 ml phate reagent. Heat to boiling to remove car- = 10 mg NH,NO,: Dissolve 10 g of reagent- bon dioxide. grade ammonium nitrate in distilled water 6.3 Add a few drops of methyl red indi- and dilute to 1,000ml. cator, and neutralize just to the yellow end- 5.4 Diammonium hydrogen phosphate point by dropwise additions of concentrated solution, 0.090 M: Dissolve 12 g of (NH,), ammonium hydroxide with constant stirring. HPO, in distilled water and dilute to 1,000 If, on addition of the indicator, a pink color ml. forms and then disappears, the water proba- bly co'ntains an excessive amount of iodide or 5.5 Ethyl ether. bromide ions. In that event, add ammonium 5.6 Magnesium nitrate reagent, 5.0 N in hydroxide, 2 or 3 drops at a time ; then add a Mg(NO,) 2, 0.5 N in HNO,: Dissolve 640 g of drop of indicator. Repeat this procedure until Mg(NO,) 2. 6H20in the minimum volume of the indicator exhibits the yellow color in- hot distilled water in a 1-litre beaker. Pour stantly upon hitting the solution. into a 1,000-mlvolumetric flask, add 32 ml of 6.4 Digest the precipitate near the boil- concentrated HNO,, dilute to volume, cool to ing point on a hotplate or steam bath for 30 room temperature, and again dilute to min ; then allow to cool to room temperature volume. and settle. 5.7 Methyl red indicator solution: Dis- 6.5 Using a small pipet connected to an solve 0.1 g of methyl red (dimethylaminoazo- aspirator, draw off as much of the super- benze-necarboxylic acid) in 250 ml of 60 per- nate as possible without disturbing the cent ethanol. precipitate. 5.8 Nitric acid, concentrated. 6.6 Transfer the precipitate to a 40- or 5.9 Uranium standard solution I, 1 ml 50-ml screw-cap Pyrex centrifuge tube. Po- = 100 pg U :See method R-1180-76. lice the beaker and the stirring rod with l 5.10 Uranium standard solution 11, 1 ml percent ammonium nitrate solution, adding = 1.0 pg U :See method R-1180-76. the washings to the centrifuge tube. Centri- fuge, discard the supernate, and add 4 or 5 5.11 Uranium standard solution 111, 1 ml ml of the 1-percent ammonium nitrate solu- =0.01 pg Pipet 10 ml of uranium stand- U: tion. Agitate the mixture in the tube to wash ard solution I1 and 5 ml of concentrated the precipitate, and again centrifuge and dis- into a 1,000 ml volumetric flask, and HNO, card the supernate. Note :In transferring the dilute to volume distilled water. with Mix precipitate to the centrifuge tube it will prob- thoroughly, and store in a Teflon bottle. ably be necessary to centrifuge and decant 6. Procedure once before the transfer can be completed. 6.7 Ovendry the precipitate in the cen- 6.1 Place 400-ml aliquots of the filtered trifuge tube for 15-20 minutes at 80°C.Raise samples in 600-ml beakers. Prepare two oven temperature to approximately 100 "C standards by addition of 8.0 ml of uranium and take to complete dryness. standard solution I11 to 400 ml of distilled 6.8 Add 8 ml of the Mg(N0,)2 reagent, water in 600-ml beakers. Also prepare a and warm gently to dissolve the precipitate. blank of 400 ml of distilled water. An addi- Use the ultrasonic generator to speed tional blank and standard (prepared by pi- dissolution. petting 50 pl of uranium standard solution I1 6.9 When cool, add 10 ml of cold ethyl directly into the platinum dish) are taken ether to the test tube, cap, and agitate vigor- METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 91 ously for at least 2 min. Allow about 15 min 1,000S(A-B) R, j pg/l of - for the two layers to separate. u= V(C-B) R, ' 6.10 Pipet off 5.0 ml of the ethyl ether where layer using a 5-ml pipet with a control, and S=rg od uranium added to prepare the place in a 125-ml Teflon dish to which ap- standards, proximately 8 drops of water have been A = mean fluorimeter reading of the added. Add another 5 ml of ether to the tube, sample disks, cap, and agitate for 2 min. Again, allow 15 B =mean fluorimeter reading of thd min for the layers to separate, then pipet an blank dish, additional 5 ml of the ethyl ether layer from G = mean fluorimeter reading of the the tube into the same Teflon dish. Use the standard disks, ultrasonic generator to break up emulsions. V=volume of the sample in milliliters, The equivalent volume transferred is thus R,= fraetional recovery of uranium ex- 7.5 ml of the original 10 ml of ether added. tracted frolm the sample,and 6.11 Allow the ethyl ether to evaporate R, =fractional recovery of uranium ex- completely in a fume hood at room tempera- tracted frorm the standard. ture. Heat gently on a hotplate to complete dryness, The fraction of uranium recovered in seri- 6.12 Transfer the sample from the Teflon al extraction of the samples and standards evaporating dish to the platinum dish used (R?,R,) is determined by equation : for fusion with a small portion of ethyl alco- hol, policing the bottom and the sides of the Teflon dish thoroughly with a small rubber policeman. Repeat the ethyl alcohol wash one more time. A third wash must be completed where using a small portion of distilled water. Com- bine with the ethyl alcohol washes immedi- v =volume of ether (ml)removed after ately. The three washes must be kept small each extraction, and enough so the combined washes do not exceed V=volume of ether (ml) in the sample the volume of the platinum fusion dish. for each extraction. (Here again the distilled water wash serves a double purpose: (1) assures complete When V is equal to 10 ml and v is equal transfer of the sample, and (2) reduces the to 5 ml (normal procedure) the R values ethyl alcohol creep up the sides of the fusion for the first, second, third, and fourth ex- dish.) Take fusion dish to dryness under a tractions are respectively 0.5, 0.75, 0.875, heat lamp. and 0.937. When the serial extraction of the sample 6.13 Carefully flame fusion dish over a is identical to the serial extraction of the burner until the dish is a dull red. standard (same number of extractions using 6.14 Add 2 g of flux and prepare fluores- the same volumes) the fractional recoveries cent disk as in method R-1180-76. cancel and the equation simplifies to: 6.15 Place the dishes in a desiccator, and 1,000S (A -B) cool for at least 30 min. Determine the fluor- pg/l of u= escence of the samples, blank, and standard V(G-B) as in method R-1180-76. 8. Report

7. Calculations Report concentrations to two significant figures above 0.10 pg/l and to one significant Concentration of uranium is calculated figure fo,rvalues below 0.10 pg/l with 0.01 from the equation : pg/l as the minimum. 92 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS 9. Precision References I Minimum detectable concentration is 0.01 Barker, F. B., Johnson, J. O., Edwards, K. W.,and Robinson, B. P.,1965, Determination of uranium pg/l. The precision of the fluorescence meth- in natural waters: U.S. Geol. Survey Water- ods for determination of uranium is gov- Supply Paper 1696-C, 25 p. Hellman, N. N., and Wolfe, M. J., 1952, Influence erned primarily by conditions in the flux and of various nitrates on the diethyl ether extrac- in the fusion operation. Reproducibility of tion of low concentrations of manium from thorium, in Production and Separation of Uran- the fluorescence from replicates standards ium-233 (ed., L. Katzin): U.S. Atomic Energy averages 215 percent. The same value is Comm. TID 5223, pt. 1. Smith, A. P.,and Grimaldi, F. S., 1954, The fluori- used for precision of sample runs except at metric determination of uranium in nonsdine concentrations below approximately 0.07 and saline waters, in Collected papers on meth- ods of analysis for uranium and thorium: U.S. pg/l where 2MDL represents the precision. Geol. Survey Bull. 1006, p. 125-131. Uranium, dissolved, isotopic ratios AI p ha spec t rorne try -c he m i cal se pa ra t ion (R-1182-76 )

Parameter and code: Uranium, dissolved, isotope ratio (dimensionless): none assigned

1. Application Uranium-238;4.195 MeV (0.77), 4.147 MeV (0.23) Uranium-235; 4.370 MeV (0.25, 4.354 MeV (0.35), The method is applicable to most fresh- (with five other energy peaks of less intensity) water and saline waters. Industrial wastes Uranium-234; 4.768 MeV (0.72), 4.717 MeV (0.28). and mine drainage may require special treat- Since surface-barrier detectors can be ob- ment. tained with resolution as fine as 0.030 MeV (30 keV) (for 450 mm2 counting area), it 2. Summary of method is possible to cleanly resolve all the uranium peaks of interest. The uranium isotopes are determined by Because of the very low radioactivity alpha spectrometry after concentration and of uranium, may be necessary to collect the separation from the bulk of the water sam- it uranium from a relatively large water sam- ple by use of the precipitation-extraction procedure described under method R-1181- ple to yield between 2.5 and 220 pg of uran- ium. Samples as large as 25 liters can be 1 76 determination of uranium. This is fol- used. lowed by an ion-exchange procedure to elimi- nate thorium, an alpha-emitting radionu- 3. Interferences clide. The final step is electrodeposition of There is no direct spectral overlap by the uranium (as uranium oxide) in a very other natural alpha-emitting nuclides. How- thin layer on a metal disk. Electrodeposition ever] if thorium-230 or protoactinium-231 conditions are critical because thick or non- are present in great excess, broadening of uniform deposition must be avoided, and in- their peaks might introduce a small error. terfering radioisotopes must be kept in the Transition metals, when present in great solution phase. Thick or nonuniform electro- excess, might be carried through the final depositions result in distortion of the alpha electrodeposition to increase the mass of the energy peaks and reduced counting effi- deposit. This would broaden the alpha energy ciency. The procedure is described in detail peaks and reduce the counting efficiency. by Edwards (1968). None of these possible adverse effects have yet been encountered practice. Alpha spectroscopy is carried out by in means of a surface-barrier detector 4. Apparatus with output to a linear amplifier and multi- 4.1 Detector, silicon surface-barrier de- channel analyzed. Readout is by mmns of tector with approximately 450 mm2sensitive an x-y plotter and electric typewriter which area and depletion depth of 60 microns or prints out counts in each energy channel. more. The alpha spectra of the uranium isotopes 4.2 Vacuum chamber. The detector is are as follows: mounted in a chamber which is evacuated

93 94 TECHNIQUES OF WATER-RESOURCES INVESTIGATIONS after the sample is inserted. This is essential 6. Procedure to minimize scattering of alphas by air. A 6.1 Determine the volume of sample re- vacuum pump and silica-gel moisture trap quired by carrying out the fluorimetriq urani- are used to maintain vacuum at 0.1 torr. um analysis by method R-1180-76 or R- 4.3 Mutticlmnnel analyzer array. The sig- 1181-76 as indicated by the nature of the nal is fed from the detector through a pre- sample. The maximum weight of uranium amplifier and linear amplifier into a multi- that can be used in the isotopic analysis is channel analyzer where the alpha energy 220 pg and the minimum is 2.5 pg. Recom- pulses are routed to their appropriate chan- mended sample volumes for a thousandfold nels in the memory. The analyzer should have range of uranium concentrations are given a minimum of 256 channels if the full resolu- in table 2. tion capability of the best silicon barrier de- tectors is to be realized. 4.4 Readout system. The memo'ry may be Table 2.-Recommended sample volumes, minimum, and reduced volumes for isotopic uranium analysis read out using a plotter, teletype unit, punched tape, magnetic tape, or other ap- Volume, in liters U concentration Reduced propriate means. The combination of x-y iUfZ/l) om2Sed ~inimum volume plotter and electric typewriter has been 0.10 -____--- 25 25 5 found to be satisfactory. .2 -__----- 25 12.5 5 4.5 Chemical-se paration apparatus. The .3 -____--- 25 8.4 5 .4 -___---- 20 6.3 4 chemical apparatus required for method R- .5 ______- 20 5.0 4 1181-76 (items 4.6 to 4.8) are used. .6 ______- 20 4.2 4 4.6 Ion-exchange columns. These are de- .8 ______20 3.2 4 signed to hold 40 ml of liquid in a funnel top 1.0 -______20 2.5 4 1.2 ______- 20 2.1 4 and to pass the solution through a 14-cm 1.5 ______20 1.7 4 length of ion-exchange resin, 1 em in 2.0 -______- 15 1.25 3 diameter. 3.0 ______10 .84 2 4.7 Electrolysis apparatus. The cell is de- signed to plate a circular deposit 2.2 em in diameter to correspond with the sensitive 12 ______5 .21 1 diameter of the alpha detector. The deposit 15 ______5 .17 1 is collected on a disk which is 20 -_ _ _ _ --- 2 .13 1 30 ______2 .084 1 clamped onto the bottom of a Teflon cylinder 50 ______2 .0,50 1 thus forming an electrolysis cup. The titani- 70 __ ___- 1 .036 1 100 _____---_ _ 1 .025 1 um disk is the cathode. The anode is a flat coil of platinum wire suspended 2.2 em above the titanium disk. Power is supplied by a 6.2 If the sample volume exceeds 1 liter, small 12-volt rectifier with voltmeter and am- evaporate on a hotplate to the reduced meter. The titanium disks are 3.2 em in volume value shown in table 2. If the reduced diamejter. volume exceeids 1 liter it is necessary to di- vide the sample into two or more 1-liter por- 5. Reagents tions. Each portion is carried through the 5.1 Electroplating reagents, NH,Cl solu- prwedure as an individual sample through tion, 2 N;acetone; and ammonia. step 6.3. Run a blank of 1-liter distilled 5.2 Ion-exchange reagents, Bio-Rad AG water through the procedure. 1-X8, 50-100 mesh or equivalent. Hydro- 6.3 Carry out the uranium extraction pro- chloric acid, 8 N and 0.1 N. cedure, method R-1181-76, steps 6.2 through 5.3 Precipitation-extraction wagents. All 6.11. If two or more 1-liter portions of one reagents required for the precipitation-ex- sample are exkracted, combine the extracts traction steps of method R-1181-76 are used. before step 6.11. METHODS FOR DETERMINATION OF RADIOACTIVE SUBSTANCES 95

6.4 Dissolve the dry residue that remains malized to the same counting time, this also from the evaporation of the ether layer in corrects f0.r background. 10 ml of 8 N HC1. The isotopic ratio is: 6.5 Prepare the ion-exchange columns by R=---,U-234 - (2-234 washing 6 g of resin in a beaker with 8 N U-238 C-238 HC1. Transfer to the columns and wash with where 8 N HCl. 6.6 Pour the solution from 6.3 into the C-234=counts under the 4.763 MeV ion-exchange tube. Allow to flow through at peak corrected for blank, and the rate of 20-30 drops per minute. Elute C--238=counts under the 4.195 MeV thorium with 50 ml of 8 N HCI and discard peak corrected for blank. this eluate. Elute uramium with 60 ml of 0.1 7.2 Determine concentration of each N HC1. uranium isotope, if desired, by applying the isotope ratio to the concentration of total 6.7 After evaporating to dryness, adding uranium as determined by fluorescence meth- HNO,,and evaporating again to eliminate the last traces of HC1, prepare the residue od R-1180-76 or R-1181-76. for electrolysis by dissolving in the electro- 8. Report lyte, 10 ml of 2 N NH,Cl. 6.8 Transfer to the electrolysis cell, and Report activity ratios less than one to two plate the uranium onto the titanium disk us- significant figures. Report activity ratios ing current of 1+- 0.1 ampere. Electrolyze for greater than one to three significant figures. 100 min. 6.9 Introduce the dry sample on the ti- 9. Precision tanium disk into the vacuum chamber, pump There are insufficient data to establish a down to 0.1 torr, and count the alpha activity reliable experimental standard deviation. of the sample for 1,000min using the energy Standard deviation based on counting statis- range 3.8-5.3 MeV. tics may be calculated using the following : 6.10 Printout the spectrum with the x-y (C-234) +B+1 (C-238) +B+1 YZ plotter and the typewriter. .-R( ((C-234) -B)’ + ((C-238) -B)’)’ where 7. Calculations B = experimentally determined blank. 7.1 Identify the uranium isotopes present Derivation of the equation may be found in on the basis of the x-y plot, using the data Edwards (1968). from the typed readout, establish the wunt under the principal alpha energy peak for Reference each isotope by summing the counts and sub- Edwards, K. W.,1968, Isotopic analysis of uranium tracting the blank. Since background for the in natural waters by alpha spectrometry: U.S. samples and the blank is the same when nor- Geol. Survey Water-Supply Paper 1696-F, 26 p.