Comparison of Nuclear Analytical Methods with Competitive Methods

IAEA-TECDOC-435

COMPARISON OF NOCLEAR ANALYTICAL METHODS WITH COMPETITIVE METHODS

PROCEEDINGS OF AN ADVISORY GROUP MEETING COMPARISON O NUCLEAF NO R ANALYTICAL METHODS WITH COMPETITIVE METHODS ORGANIZEE TH Y DB INTERNATIONAL ATOMIC ENERGY AGENCY AND HELD IN OAK RIDGE, 3-7 OCTOBER 1986

A TECHNICAL DOCUMENT ISSUED BY THE INTERNATIONAL ATOMIC ENERGY AGENCY, VIENNA, 1987 COMPARISO NUCLEAF NO R ANALYTICAL METHODS WITH COMPETITIVE METHODS IAEA, VIENNA, 1987 IAEA-TECDOC-435

Printe IAEe th Austri n Ai y d b a October 1987 PLEASE AWARE B E THAT MISSINE TH ALF GLO PAGE THIN SI S DOCUMENT WERE ORIGINALLY BLANK The IAEA does not normally maintain stocks of reports in this series. However, microfiche copies of these reports can be obtained from

IN IS Clearinghouse International Atomic Energy Agency Wagramerstrasse 5 P.O. Box 100 A-1400 Vienna, Austria

Orders shoul accompaniee db prepaymeny db f Austriao t n Schillings 100, for e for e chequa th f m th IAEmf o n i o n i r Aeo microfiche service coupons which may be ordered separately from the INIS Clearinghouse. FOREWORD f nucleao e us r e analyticaTh l techniques, especially neutron activation analysis, already 0 yeahavd 5 histor ol a re y behind them. Nuclear methods provided means for reliable trace element analysis in a time when few other trace element analysis methods were available. Therefore the role of nuclear analytical techniques in developing science and technology, certainly have been important.

Today several sensitive and accurate, non-nuclear trace element analytical technique e availabl ar w methods ne d e continuouslan ar se y developed. The Agency is supporting the development of nuclear analytical laboratories in its member states. In order to be able to advise the developing countries which methods to use in different applications, it is importan o kno t e presen wth t statu d developmenan s t trend f nucleao s r analytical methods. Whic e theiar h r benefits, drawback d recommendean s d fields of application, compared with other, non-nuclear techniques.

In order to get an answer to these questions the Agency convened an Advisory Group Meetin k RidgeOa n i g, Tennessee 7 Octobe3- , r 1986. This volume is the outcome of the presentations and discussions of the meeting.

e Agenc e expertTh th s gratefu i yl o contribute al wh s o t l d paperd an s too e kdiscussions th par n i te chairmath k Oa o . t ,H.He Dr nth .o t Rosd an s Ridge National Laboratory, which hosted the meeting in such an efficient and successful way. EDITORIAL NOTE

In preparing this materialpress,the International the for staff of Atomic Energv Agenc\ have mounted and paginated the original manuscripts as submitted by the authors and given some attention presentationthe to The views expressed in the papers, the statements made and the general st) le adopted are the responsibility of the named authors The views do not necessarily reflect those of the govern- ments of the Member States or organizations under whose auspices the manuscripts were produced thisin The bookuse of particular designations countriesof territoriesor does impl\not an\ judgement b) the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities institutions delimitationand the of or theirof boundaries The mention specificof companies theirof or products brandor names docs unpl\not an\ endorsement 01 recommendation on the part of the IAEA Authors themselvesare responsible obtainingfor necessar)the permission reproduceto copyright material from other sources CONTENTS

Introduction ...... 7

Statistical evaluation of recorded knowledge in nuclear and other instrumental analytical techniques ...... 9 . T. Braun Discussion ...... 6 3 . f scientometrico e Us asseso t s s nuclea othed an r r analytical methods ...... 7 3 . W.S. Lyon Comments on the establishment of a methodology for the metrological characterization of analytical methods ...... 49 E.H. Klehr, J. Tölgyessy Discussion ...... 7 5 . Plutonium isotopic measurements by gamma-ray spectrometry versus mass spectrometry (abstract only) ...... 9 5 . R.L. Mayer, D.A. Rakel Discussion ...... 0 6 . Uranium determination by fission track counting ...... 61 F.F. Dyer Discussion ...... 6 6 . Applications of Cerenkov counting: A useful technique for environmental radio-analyses (abstract only) ...... 7 6 . J.S. Eldridge Discussion ...... 68 Comparison of isotope dilution analysis with other analytical techniques ...... 69 J. Tölgyessy, E.H. Klehr Discussion ...... 82 The contribution of radiotracers to the development of trace element analysis ...... 83 V. Krivan Discussion ...... 8 8 . Applicatio isotope th f no e exchange metho f analysido speciative th o t s e determinatiof no phosphoru environmentan i s l samples ...... 9 8 . N. Ikeda Discussion ...... 103 Resin bead methodolog s applieya fueo dt l burn-u fissild pan e inventories ...... 5 10 . D.H. Smith, R.L. Walker, J.A. Carter Discussion ...... 6 11 . A survey of immunoassay techniques for biological analysis ...... 117 C.A. Burtis Discussion ...... 127 e rol f INATh eo s compareAa conventionao dt l method f analysio s r geologicafo s l sample n Canadi s a ...... 9 12 . E.L. Hoffinan Discussion ...... 141 Determination of minor and trace elements in diets using neutron activation, inductively coupled plasma atomic emission and atomic absorption spectrometry (abstract only) ...... 143 R. Schelen- Discussion ...... 4 14 . Presen futurd an t e prospect r neutrofo s n activation analysis compareo dt 7 14 other methods available G Revel Comparison of 14 MeV neutron activation analysis and competitive methods for determination of oxygen, nitrogen, silicon, fluorine and other elements 163 BildW R 8 17 Discussion Nuclear analytical methodElectricaT NT t a s l Communications Laboratories Substoichometnc analysis and its applications 179 T Shigematsu Nuclea7 18 r analytical method standardn i s s certification R M Lindstrom Discussion 193 Analysi f difficulo s t materials Compariso f nucleano d non-nucleaan 5 r 19 r methods C J Pickford, J S Hislop Discussion 200 Trace element anal} sis in biological materials A comparison of neutron activation analysis with other techniques, especially atomic spectroscop} 201 P Schräme! The development of activation analysis in China 213 T Xinhua 7 21 Discussion Neutron activation analf geologicao s }si l sample fren i s e competitio— n 9 21 A case histor} from Finland R Rosenberg Lipponen,M L Vanska 0 23 Discussion

1 23 Summar Conclusiond }an s

7 23 Lis f Participanto t s INTRODUCTION

An Advisory Group Meeting (AGM) was held on the 3rd, 4th, 6th, and 7th of Octobe rk Ridge Oa 198 n i 6, Tennesse s attendewa d an y escientistb d s from over ten countries in Europe, Asia, and North America. The purpose of this AGM o examint s e rol wa f th nucleaeo e r method f analysio s s vis-a-vis non-nuclear techniques. Our specific objectives were to evaluate critical parameters associated with various analytical methods, determine the range of application methodse th f o , sugges w criticaho t l analytical technolog e effectivelb n ca y y transferred to developing nations, and most important, counsel the Agency on possible program initiatives that they might establish to encourage such technology transfer. Eac f theso h e objectives f courseo , , encompassea d variet f relateo y d questions r exampleFo . , problems related o analyticat l reference materials, personnel training n evaluatioa d an , f strengtho n d an s deficiencies of key analytical methods were included subjects of discussion. It is clear that the range of topics considered by the AGM was very broad and t verye y results oriented.

e meetine formaTh th f o tg consiste f essentiallo d o parts tw ye forma th : l presentation d comments an se specia th d an ,l sessions devote o discussionst d . The schedule was kept very flexible to encourage an active exchange of ideas without having to worry about the "tyranny of the clock". Conclusions were first elaborated by four groups of individual session members led by a rapporteur and were later modified and/or expanded during a general meeting of the delegates. Finally e explicith , t suggested directive o the'Agenct s y were prepare meetina e sessio t th a d f o gn rapporteur e conferencth d an s e chairman.

This volume contains the texts of the formal presentations, the discussion that followed each, and the summary and conclusions of the AGM. It is also evident that much of what is said extends far beyond the interests f juse developino th t g countries; chemical analysi truls i s a globay l concern with global problems.

I would like to express my sincere appreciation to W. D. Shults and W. R. Laing of Oak Ridge National Laboratory and R. Rosenberg and V. Vasilyev of the IAEA for their significant contributions in helping plan and organize this meeting. Finally, I would like to thank all of the delegates to the AGM. Their interesting presentations and their open and often spirited discussions y factorarke e whan i s mos a believI t e tb successfuo t e l outcome.

Harley H. Ross, AGH Chairman Oak Ridge National Laboratory STATISTICAL EVALUATIO RECORDEF NO D KNOWLEDGE IN NUCLEAR AND OTHER INSTRUMENTAL ANALYTICAL TECHNIQUES

. BRAUT N Institut f Inorganieo Analyticad can l Chemistry, L. Eötvös University, Budapest, Hungary

Abstract

The main points addressed in this study are the following. Statistical distribution patternf so published literatur n instrumentao e l analytical techniques 1981-1984. The structure of scientific literatures and heuristics for identifying active specialtie emergind an s t spoho gt research areas in instrumental analytical techniques. Growth and growth rates of the literature in some of the identified hot research areas. Quality and quantity in instrumental analytical research output.

There are basically two ways of assessing or evaluating the progrès, trends, advances, growth, etc., of a given field r subfielo f scienceo d :

1. Peer evaluation 2. Statistical evaluatio f recordeo n d knowledge /subject literature/ When the first of the abovementioned ways is used the paper, report, etc., communicating the results of the assessment begins usually wit sentenca h e followinth f o e g type: "This paper /report/ provides my personal opinion on the status /the trends, progress, etc./ of ...... its present strengths, and its weaknesses. I also speculate on what the future will provid d indicatan e e wher e researcth e h opportunities appea o be.t r " Mos f thito s typ f assessmeno e t papers, writte wely nb l recognised individuals e base,ar n commoo d n shared insights, casual impressions and estimations, but they seem in general to lac a solik d empirical base. e evaluatioTh n use n thii d s pape e seconr th opte r dfo d abovementioned way i.e. it is using the statistical evaluation e literaturth f o f analyticao e l chemistrye th e goa Th .f o l effor s bee tha o char t n t comparativel whaf o ys studie twa d in the field of nuclear and other instrumental analytical techniques o follot , e wtrendline th som maie f o eth n f o stopic s e whertse o e this typ f researco e e past th s bee ha hn , i n currentl e d headingseemb an o t , s is y.

The reader intereste e detailth n f i thido s s type of assessment s recommendei s o consult d t some previsous 1 c: publications. J In what follow e havsw e collecte quita d e comprehensive amount of empirical data on the research and publication activity in the field of nuclear and other instrumental analytical techniques e date collate.Th ar a n tablei d d chartan s s with very brief interpretation. That's because the reader is invited and encourage o examint d e table e whath ed chartse an tsd an s additional comments, conclusions he/she can observe. Tabl e distributioshow1 e th e dat n th so a f useo nf o s different instrumental analytical technique n analyticai s l research papers published durin e 1981-198th g 4 periode .Th databas r thifo es analysie 1981-198th s swa 4 volumee th f o s Analytical Abstracts. A set of 15 elements has been selected /see the list of elements in Table 2/ and for each element the individua f instrumentao e us l l analytical technique s beeha s n

10 Tabl1 e

Frequency distribution of the uses of instrumental techniques in 1981-1984 research papers World Total

Technique usef o s. No

atomic-absorption spectrophotometry 1648 24.1 spectrophotometry 1271 18.5 emission spectrornetry 1005 14.7 neutron-activation analysis 585 8.5 6 6. 2 45 x-ray fluorescence 2 2. 9 14 thin-layer chromatography high performanc7 1. e liqui 7 11 d chromatography 7 1. 6 atomic-emissio11 n spectrometry 4 1. 4 9 polarography 2 1. 1 8 ionchrornatography differential pulse polarography 61 0.9 ionseiective electrode 59 0.3 flow injection analysis 57 0.8 8 0. 6 5 kinetic methods 8 0. 2 5 fluonmetry anodic stripping voltammetry 49 0.7 voltarnmetry 47 0.7 inductively coupled plasma 46 0.7 proton induced x-ray emission 45 0.7 atomic-fluorescence spectrornetry 42 0.6 particle induced x-ray emission 40 0.6 coulometry 32 0.5 mass spectrometry 32 0.5 differential pulse cathodic stripping vol tanmietry 31 0.5 4 0. 9 2 amperometry stripping potentiornetry 26 0.4 gas-chromatography 25 0.4 osci1 Iopolarography 23 0.3 2 0. 7 1 chemiluminescence isotope dilution 17 0.2 2 0. 4 1 Auger electron spectrometry electron microprobe 12 0. 21 liquid chromatography 12 0.2 1 1 0.2: conductometry 1 0. 9 polarograph. a.c y mol. emission cavity spectrometry 9 0.1 1 0. 9 paper chromatography 1 0. 8 photoâcoustic spectrometry proton activation Q 0.1 stripping(chrono)voltammetry 8 0.1 electro1 n 0. spir paramagnetic(o n 7 ) resonance f. a. n.e. s. 7 0.1 inverse voltarnmetry 7 0.1 1 0. 7 pulse polarography

Miscellaneous 420 6.1

Total 6852 100.0

11 Tabl2 e

Percentage distributio e use th f mergeo sf o n d instrumental analytical techniques in 1981-1984 research papers

World total

Technique* ———————————————— Element ———————————————————————— Total I Ag As Au Bi Co Cr Cu Fe Ga Hg Mn Mo Sb U V I 17.) (7.) (7.) <7.) (7.) (7.) (7.) m (7.) 17.) (7.) (7.) (.7.1 (X) (7.) (7.) I OPT 53.1 70.4 53.1 67.9 62.9 70.2 66.3 70.8 71.6 65.9 73.2 70.1 59.5 47.2 74.7 66.9 NUCL 14.1 15.1 25. 3 13.93. 6 14. 2 13.08. 5 11.9 12.9 16.2 12.6 21.6 29.1 11.1 13.8 ELEC 15.7 7.8 1Û.O 15.9 9.5 7.4 14.6 5.6 12.8 9.8 3.9 9.6 11.4 10.9 7.2 9.8 8 5. 6 3. 8 6. 4 3. 8 4. 1 4. 1 9. 8 1. 2 5. 8 6. 3 10. 9. 8 05. 7 2. 2 4. 5 CHRO3. « MISC 7.7 2.5 2.3 3.7 4.1 2.6 4.1 4.9 1.0 2.3 2.6 3.0 4.2 2.9 3.4 3.7

0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 10 0 1010 0

r categori:atioFo 1 n into merged group e Appendise s 2 x

optical: T OP i 2 spectroscopical; NUC nuclear: L ; ELE electrochemica: C CHRO; l Chromatographie: M ; MISC : miscellaneous

counted in the subject index of each volume of the Analytical Abstracts. As such it has to be accentuated that the counting doesn' te numbe refeth f o papero rt r n instrumentao s l techniques but to the number of uses of these techniques. That's-why e.g. f mano one y eus instrumenta papee n refeth ca ro t r l techniques. e differenTh t instrumental analytical techniques taken into account wert selecteno e d i.e. define. Theus y y werb d e considere e countinth n i d s thega y were define d mentionean d d e subjecith n t index classification e Analyticasysteth f o m l Abstracts database. The 6852 uses of different instrumental analytical techniques were distributed between a total of 143 distinct techr.i<~ -ics /see list of techniques in appendix I/. As visible, in tab!.- I we did collect only those 44 methods t leas a whice researc d th 7 uset ha hn i s h papers publishen i d

12 the 1981-1984 period. The data show a very skewed distribution e use oth f differeno s t techniques. A special type of cumulative-frequency graph known as the "Lorenz curve" has been used effectively to portray distribution data e distributioth , e.g. wealtf no d incoman h e in relation to certain segments of the population, the productivity of farms in terms of cumulative proportions of e farmsdistributioth r ,o f retaio n l sale s relatea s o t d various groupings of stores. Figure 1 represents the Lorenz curve for the frequency distribution of instrumental analytical techniques.

Spectropho turnstry Emission spectrometry——————- Heutron activation analysis A-ray fluurescjznce__spectrorretri! Tnir-layer chrorra teg räch j hign perfc^r-ance liquid cnro^otc

- Folarograpnj --Ion chrorrato$"aphj ——— - Differential p-ulse polarz - Ion selective electrodes - FL.OU injection analysis

10 20-30—40—50-^60-70^8 0 JOO9 - 0— ——————————— iterp^tnaf-can^tffd- uses — -___„____'_ ____L

— H —-— ' -_ i - - i

1 Lorenz-typ - c ri e cuulative frequency clistijLutiof o e rv i c n

e use th f instrumentao s l analytical techriruee th n i s 1931-1984 period

13 As shown, the distribution in uses of these methods is fairly "undemocratic" i.e. about 10 % of the methods account e usesth .f r approxo fo Spectrometri% 5 8 . c method e higar s h e rankinith n g with atomic absorption spectrometr n stroni y g first positioo nucleatw t bu rn analytical techniques /neutron activation analysis and XRF methods/ are also highly quoted. Table 2 shows the distribution of the uses of instrumental analytical techniques 5 elementaccordin1 e th so t g taken

into account e differen th e dat n Th .o a t technique e mergear s d here into 5 comprehensive groups. /Optical i.e. spectrometrie, nuclear, electrochemical, Chromatographie and miscellaneous; for merging into groups see appendix 2.1

en a. oU- 10' o

D r „ • 0 9 y u

7l T> td «l 82 «3 Ï«

Figur 2 eCumulativ e worle growtth df o hliteratur n floeo w injection

analysis /tD: exponential doubling time, i.e. growth rate/ '2| : dates of major conferences

14 As visible, nuclear analytical techniques are in strong second position after the spectroscopic ones. It is interesting to compare the ranking data of table 2 r previouou f o e resulto sth tw ranking o f t e differeno sth f so t instrumental analytical techniques /Figure 2/. ' We see nuclear analytical techniques ranked fourt e previou th botn i hf o h s investigations.

Tables 3-14 presen e geographicath t l distributiof no the uses of the different instrumental analytical techniques /see appendix 3 for defining the different geographic regions/.

Tabl3 e

Frequency distributio e use th f instrumentao sf o n l techniques in 1983-1984 research papers

USA & Canada

Technique No. of uses

atomic-absorptio 7 14 n 26.spectrophototaetr8 y emission spectroaetry 84 15.3 neutron-activation analysis 73 13.3 4 8. 6 4 x-ray fluorescence 4 6. 5 3 atomic-emission spectroinetry 8 3. 1 2 spectrophotometry 3 3. 8 1 inductively coupled plasma Mass spectronetry 16 2.9 high performanc2 2. e liqui 2 1 d chromatography 2 2. 2 1 ionchromatography electron microprobe 9 1.6 flow injection analysis 8 1.5 1 1. differentia 6 l pulse polarography 3 0. 5 liquid chromatography 7 0. 4 anodic stripping voltamtnetry 5 0. 3 gas-chromatography coulonetry C3 0.5 differential pulse catliodic stripping vol tasunetry 3 0.5 4 0. 2 ions«l ective electrode isotope dilution 2 0.4 Auger electron spectroifietry 2 0.4 4 0. 2 thin-layer chromatography atottic-f luor*scence spec t r omet r y 2 0.4 2 0. l aihperofnetry polarography l 0.2 fluor i i»*t r y l 0.2 2 0. l stripping(chrono)voltammetry electron2 0. spir paramagnetic(o n 1 ) resonance

Miscellaneous 29 5.3

9 54 100.0 Total

15 Tabl4 e

Frequency distributio e use th f instrumentao sf o n l techniques in 1983-1984 research papers

United Kingdom

. No Technique of uses X

at omi c -absor p 1 1 on sp ec t r ophot omet r y 47 26.7 emission spectrometry 30 18.3 spectrophotoirietry 17 10.4 flow injection analysis 11 6.7 atomic -emission spectrometry 10 6.1 x-ray fluorescence 10 6. 1 high performance liquid Chromatograph/ 9 5.5 inductively coupled plasma S 5.5 ionchromatography 2 1.2 mass spectrometry 2 1.2 kinetic methods 1 0.6 f luoriraetry 1 0.6 neutron-activation analysis 1 0.6 atofuic-oncc es e r spectronietruo 1 f y 1 0.6 particle induced x-ray emission 1 0.6 pol arography 1 0.6 differential pulse cathodic strippin tatumetrl vu g y 1 0.6 mol. emission cavity spectroinetry 1 0.6

Misce arieou1 l s 9 5.5

Total 164 100.0

Table 5

Frequency distribution of the uses of instrumental techniques in 1983-1984 research papers

Fed. Rep. Germany

Technique Ni. of uses

atomic-absorptio 5 6 n 29.spectrophctonietr7 y «mission spectrometry 24 11.0 x-ray fluorescence 19 8.7 5 5. 2 1 neutron-activation analysis 5 5. 2 1 vol tanitc.etry 6 4. 0 1 gas-chromatography high performanc1 4. e liqui 9 d chromatography spectrophotometry 7 3.2 2 3. particl 7 e induced x-ray emission 2 3. 7 differential puls arographl po e y Auger elertron spectrometry 5 2.3 differential pulse cathodic strippin8 1. g voltammetr 4 y proton activation 4 1.8 lonstrl ecti v«? electrode 3 1.4 4 1. 3 mass spectronitftry 9 0. 2 amperometry atomic-etui ssion spectrometry 2 0.9 isotope dilution 2 0.9 anodic stripping vol tanirnetry 2 0.9 9 0. 2 thin-layer chromatography 5 0. 1 inductively coupled plasn.a kinetic methods 1 0.5 5 0. 1 couloiuetry 5 0. 1 pol arography 5 0. 1 photoacoustic spectrometry lonchroinatography 1 0.5

5 5. 2 1 laneouI e Mse i s

9 21 100.O Total

16 TabLe 6

Frequency distributio e use th f instrumentao sf o n l techniques n 1983-198i 4 research papers France

Technique No. of uses 7.

emission spictroiiietry 9 22.0 at o«ic -absorption sptctrophotometry 8 19.5 x-ray fluorescence 4 9.8 proton activation 3 7.3 neutron-activation analysis 2 4.9 spec t rophot omet r y 2 4.9 differential pul so polarography 1 2.4 f luor l met r y 1 2.4 anodic stripping voltammetry 2.4 inductively coupled plasma 2.4 particle induced x-ray emission 2.4 mass spe omer c t y tr 2.4 electron micrope rob 2.4 atomic -emission spectrornetry 2.4

Mi seel 1 ancous 5 12.2

Total 41 IOO.O

Table 7

Frequency distribution of the uses of instrumental techniques in 1983-1984 research papers USSR

Technique Nof useo . s

4 12 13.0 sprctropliotoriietry 6 8 l&.Oemission spectrometry atomic-absorption spectrophotometry 73 13.5 x-ray fluorescence 40 7.4 1 7. 8 3 neutron-activation analysis 2 2. 2 1 lonselective electrode polarography 11 2.O lonchromatography 10 1.9 voltamroetry 10 l.S 3 1. 7 kinetic methods thin-layer chromatography 7 1.3 1 1. atomic-fluorescenc 6 e spectrometry 1 1. particl 6 e induced x-ray emission 1 1. & amperometry differential pulse cathodl9 0. c strippin 5 tammetrl vo g y f luorirnetry 5 0.9 7 0. 4 anodic stripping voltammetry coulometry 4 0.7 a. c. pilarography 4 0.7 inverse vol tarumetry 4 0.7 cheitl luminescence 3 0.6 £ O. 3 »ass spectrometry G O. 3 paper chroraatography photoacoustic spectrometry 3 0.6 stripping(chrono)voltammetry 3 O.C 6 0. 3 osciIlopolarography electron4 spi0. r paramagnetic(o n 2 ) resonanre high performance liquid chromatography 2 0.4 ^ 0. 1 isotope dilution liquid chromatography 1 0.2 inductively coupled plasma 1 0.2 stripping potentlometry 1 0.2

5 3. 1 5 Miscellaneous

9 53 100.0 Total

17 Table 8 Frequency distribution of the uses of instrumental techniques in 1983-1984 research papers Japan

Technique No. of uses

atomic-absorption spectrophotowetry 95 29. 0 spectrophotowetry 90 26.5 high performanc5 6. e liqui 2 2 d chromatography 9 3. 3 1 xy fluorescenc—ra e neutron-activation analysis 10 2. S emission spectrometry 9 2.7 7 2. 9 atomic-emission spectrometry 1 2. 7 flou injection analysis lonselective electrode 6 l.B lonchronatography 6 l.B 2 1. 4 atomic-f luorescence spe omer c t y tr particle induced x-ray emission 4 1.2 4 1.2.photoacous c tspei c trowetry gas-chromatography 3 0.9 anodic stripping vol tai«uetry 2 0.6 6 0. 2 inductively coupled plasma 6 0. differentia 2 l pulse polarography 6 0. 2 polarography mass spectronetry 2 o. £ thin-layer chrowatography 2 0.6 6 0. 2 chem uminescencI i e f I uoritnetry 2 0.£ differential pulse cathodic stripping vol tan.inetry 1 0.3 3 Q. l ainperowetry 3 0. 1 stripping potent lor.ietry 3 0. 1 kinetic methods coulofrietry 1 o. 3 3 o. I isotope dilution Auger electron spectro«ctry 1 0.3 3 0. l liquid Chromatograph/ vo I tammet r y l 0.3 3 0. 1 proton activation

31 9.1

Total 339 100.0

Table 9 Frequency distributio e use th f instrumentao sf o n l techniques in 1983-1984 research papers Asia

Technique f useo . s No spectrophototnetry 222 41.3 atomic-absorption spectrophotometry 63 11.7 3 3. 8 1 thin-layer chroraatoqraphy «mission spectrowetry 17 3.2 x-ray fluorescence 16 3.O amperometry 16 3.0 8 2. 5 1 neutron-activation analysis polarography 12 2.2 lonselective electrode 11 2.O anodic stripping vol tauuaetry 11 2.0 lonchromatography 11 2.0 9 1. 0 1 atomic-emission spectrowetry inductively coupled plasma 9 1.7 3 1. 7 osciIlopolarography differential pulse cathodic stripping vol tanuuetry 5 0.0 fluorimetry 3 0.6 3 O.Eflow .injection analysis paper chrcxaatography 2 '1.4 kinetic methods 0.2 vol tamnietry 0.2 2 0. differential pulse polarography couloxetry 0.2 2 . 0 r ipping(cnroiiost tammetrl )vo y

2 B 15.2 Miscellaneous

Total 538 100.0

18 Tabl0 1 e Frequency distribution of the uses of instrumental techniques in 1983-1984 research papers Latin America

Technique No. of uses X

atomic-absorption spectrophotometry 10 30.3 spectrophotometry 8 24.2 neutron-activation analysis 2 6. 1 x-ray fluorescence 2 6.1 differential pulse polarography 1 3.0 ionsel ective electrode 1 3.0 flow injection analysis 1 3.0 kinetic methods 1 3.0 pulse polarography 1 3.O

Miscel 1 aneous £ 18.2

Total 33 100.0

Tabl1 1 e

Frequency distribution of the uses of instrumental techniques n 1983-198i 4 research papers Eastern Europe

X f useo . s No Technique

atoroic-absorptio 7 8 n spectrophotowetr23.C y 1 . 14 2 5 spectrophotometry emission spectrometry 47 12.8 S . 7 9 2 x-ray fluorescence neutron-activat ion analysis 26 7.1 Ê 4. 7 1 lonchromatography 6 4. 7 1 anodic stripping voltawmetry 0 3. 1 1 thin-layer chromatography polarography 11 3.0 lonselective electrode 5 1.4 4 1. 5 voltaroraetry 4 1. 5 . polarographc . a y stripping potent lotnetry 4 1.1 differential pulse polarography 4 1.1 Q O. 3 atomic-emission spectrometry conductoroetry 3 0.8 differential pulse cathodic stripping vol tatntuetry 3 O.Q gas-chromatography 2 0.5 liquid chrodiatography 2 0.5 particle induced x-ray etmssion 2 0.5 coulometry 2 0.5 5 0. 2 stripping(chrono)vol tammetry 3 0. 1 isotope dilution 3 0. 1 amperometry inductively coupled plasn.a 1 0.3 flow injection analysis 1 0.3 chemiluminescence 1 0.3

5 6. 4 2 Miscellaneous Total 368 100.O

19 Table 12

Frequency distributio e use th f instrumentao sf o n l techniques in 1983-1984 research papers

North Afric i Neaa r East

Technique No. of uses

spectre-photometry 8 20. 0 high performance liquid chromatography 5 12.5 atom ic -ab sorption spectrophotornetry 4 10.0 atomic-emission spectrometry 4 10.0 x-ray fluorescence 3 7.5 neutron-activation analysis 2 5.0 particle induced x-ray emission 2 5.0 1 iquid chromatography 2. 5.0 ionchromatography \ 2.5 emission spectrometry 1 2.5 pol arography 1 2.5 paper chromatography 1 2.5

i seeM aneou1 l s 6 15.O

Total 40 100.0

Table 13

Frequency distribution of the uses of instrumental techniques in 1983-1984 research papers

Res f Africo t a

Technique f useo s. No

spectrophotonietry atomic-absorption spectrophotornetry Miscellaneous 1 11.1 Total '3 100. 0

20 Tabl4 1 e Frequency distribution of the uses of instrumental techniques in 1983-1984 research papers Rest of Developed Countries

Technique No. of uses

atomic-absorption spectrophotowetry 150 26.5 6 9 17.0 spectrophotometry emission spectrooetry 52 9.2 2 6. 5 3 neutron-activation analysis 7 3. 1 2 x-ray fluorescence fluorimetry 17 3.0 particle induced x-ray emission 17 3.O atomic-emission spectrornetry 12 2.1 ionchronatography 11 1.9 9 1. differentia 1 1 l puts* polarography kinetic methods 9 1.6 differential pulse cathodic stripping voltamraetry 9 1.6 anodic stripping voltammetry 8 1.4 high performance liquid chromatography 8 1.4 4 1. 8 lonselective electrode stripping potentiometry 8 1.4 gas-chromatography 7 1.2 polarography 5 0.9 thin-layer chrutnatography 5 0.9 9 0. 5 »ass spectrotnetry 7 0. 4 inductively coupled plasma 5 0. 3 voltammetry conductometry 3 0.5 5 0. 3 paper chromatography 4 0. 2 coulometry 4 0. atomic-fluorescenc 2 e spectrornetry 4 0. 2 cheroiIuminescence 4 0. 2 electron »iicroprobe flow injection analysis 2 0.4 4 mol0. . emissio 2 n cavity sptctror.irtry a*perofnetry 2 0.4 2 0. 1 liquid chromatography isotope dilution 1 0.2 2 0. 1 stnpping(chrono)vol tarnroetry pulse polarography 1 0.1

Miscellaneous 41 7.2

Total 566 100.0

That scienc mosaia s i ef specialties o c a t no d ,an unified whole, either socially or intellectually, is a frequently made assumption. Most scientists have intuitive notions abou e subdivisionth t f theio s r fieldso observern t ,bu , however broadly trained n gaica , n overall perspective th n o e scientific mosaic s generalli t I . y supposed that specialties are effectively organized as communication systems around certain key individuals and documents in science. Recently,

21 computer-based techniques have been devise o identift d y clusters of highly interactive primary journal papers. It is contended that these clusters represen e scientifith t c specialties which currently exhibit high levels of research activity. Early studies established the feasibility of examining specialties in terms of citation data. Frequently cited papers e employear basis a d c unit r analysisfo s . This choic s basei e d on evidence that the onset of rapid specialty growth is accompanied by the emergence of key papers which are quickly and frequently cited; co-citation is used as a measure of the association between pair f frequently-citeo s d papers e strengtTh . h of co-citation is defined as the number of times two papers have been cited together providet i : naturaa s d quantitativan l e way to group or cluster the cited papers. Co-citation identifies relationships between papers which are regarded as important by authore specialtyth n i s . The strengt f co-citatioo h n betwee o e paperb tw nn ca s readily determined from the Science Citation Index^ . o papertw s e locatee i indes th Eacth d thei f an n xo hi d r lists f citino g paper e scannedar s e numbeTh . f identicao r l citing items /i.e. newer papers citing both the earlier papers/ furnishes

a quantitative measure strenghth f o ef co-citatio o t n between the two papers, or of the extent to which two items of earlier literatur e citear e d togethe e lateth ry b r literature. Therefore, co-citatio a relationshi s i n p whic s recognizei h d an d maintaine y currenb d t researchers. One way to think of a co-citation relationship is as

a matri f similarito x y coefficients e morTh . e ofte o articletw n s e citear d together e higheth , e probabilitth r y that thee ar y see y researchernb e fielth s similaa n di s o eact r h othen i r

intellectual focus. Such similarity matrices can be analyzed by a technique known as multidimensional scaling, an iterative procedure that can produce a plot of the set of highly cited o paperdimensiontw n i s s such thae highlth t y co-cited papers are placed near to each other and the little co-cited papers e placear d farther apart e relativ.Th e distances between papers in the plane should then be an indication of the degree of their intellectual relatedness o producT . e relatedness diagrams, each pape wert i s treate i rGaussiaa ef i s a d n hill whose heighs i t the number of times it was cited /its visibility/ and whose width is arbitrarily set at a value convenient to produce attractive pictures e Gaussian.Th e addear s d togethed an r resulting plo f hill o td valley an s s give w muc a senssho h f o e activity is going on in different parts of the "specialty space". The abovementioned general approach has been used recently to test a co-citation based clustering approach for identifying emergent areas /clusters/ of growth in science. This study gave a list of growth topics or potential "hot spots" in science. Tabl 5 show1 e a selectios e "hoth tf no spots" enumerated in the abovementioned study which we considered to be of some relevance to the instrumental analytical techniques and general analytical chemistry field. A quantitative illustration that most of these heuristically selected specialtie e reallar s y "hot spotss i " see n Figurs visibli nA . e 2 e researcgrowte th th e f o h h literature n floo w injection analysis /FIA/ durin e 1975-8th g 5 period s beeha n exponential, wit n averaga h e growth rate /doubling time, t/ of 1.3 years. There are however also other specialties in the

23 Tabl5 1 e n Emerging Specialty "Hot Spot" Clusters /Analytical Chemistr Instrumentad an y l Techniques/

- Chemically Modified Polymer Electrodes - Pollution of Aquatic Environments

- Atomic Fluorescence Spectrometry - Sulphur Compounds in the Atmosphere - Field Desorption Mass Spectrometry - Technique f Photoacoustio s c Spectroscopy - Mass Spectrometry

- Atmospheric Chemistry r PollutioAi : n

- Analytical Electrochemistry: Methodology

and Application of Dynamic Techniques

- Continous-Flow Injection Analysis

instrumental analytical field which show similar fast growth

rates s examplA . e presen4 Figure d e an growt th 3 ts h ratef o s n chromatographio f activatioo d an y nn o analysis e se e W .

4 thaFiguree growtd th t an year2 3 2. hs d sratean s3 1. wer f o e

respectively which can be considered tremendously fast in e abovementioneth l al d cases a comparison Figur s o A .e 4 ese e w n

e growtth h rate e tota/t,th f /l o literatur analytican eo l chemistr f beino y g 13.9 years /doubling time/ theso D . e exponential /fast/ growth rates really reflect growth of knowledg e respectivth n i e e instrumental specialties t wouldI ? ,

provided two assumptions were correct: 1. all knowledge obtained

by instrumental analytical researchers would be included in

the literature, and 2. every paper would contain either an

equa r knowo l n proportio f thio n s knowledge. Obviously neither

24 1975 1976 1977 1978 «979 1930 Year

Figure 3 Cumulative growth of. the world literature on ion chromatography. Curv : journaA e l papers; Curv: B e journa conferencg l e papers; Curv : cumulativC e e growth

e numbeth of f authoro r s /T^: exponential doubling time,

i.e. growth rate/

Figur 4 eCumulativ e e growtworlth f do h literatur f activatioo e n

analysis and analytical chemistry /T,; exponential doubling time, i.e. growth rate/

25 is true. Indeed, growth of knowledge in the instrumental analytical fiel rathea / d r abstract concept s differeni / t from growth of its literature. Moravcsik 1 9 distinguishes three relevant quantities when studying growth of science: scientific activity, scientific productivity and scientific progress. These he illustrates by analogy. A man desires to go from one placdensa n i ee forres o anothet e fivon r e miles away. Activity corresponds to the amount of work done, for example, thrashin ge undergrowthabouth n i t , blazing trails, exploring, etc. Productivity correspond e amounth y whicho b t s , as a result of all these activities, he advances toward his goal. Progres e srati th the f productivits o i n e totath o lt ytask .

Scientific activit s concernei y d wite consumptioth h n of input resource s relatei d o an factorst d s suc s numbea h f o r scientists involved and research expenditures. Scientific productivity extene referth whico o t t s h this consumptiof no resources create boda s f o y scientific results, which usually appea s publishea r d research papers. Scientific progress refers extene tth o whico t h scientific activit measure- y d through scientific productivity - actually results in substantial contributions to scientific knowledge. Rescher define quantitya s sucX , h a totatha n i tl population of papers /p/ at time t, p/t/ there will be p /t/ paper eact a s level\ h least a s follows a ,, t1 routine= A : ; A = 0.75, at least significant; A = 0.50, at least important; = 0.25X ,t leasa t very important /see also Figur. 5/ e Accordin e abovementionedth o t g e literaturth , A f o e quality grows wit a doublinh g tim f t,/Ao e .r worl fo Wit, dt- h analytical chemistry literature at about 13 years /see Figure 4

26 X-l x-i x-o

routine significant important very important first-rale

QUANTITE TH QUALITF YO Y

the numbe f itemo r s —out of the (otal population of QUANTITY - whic- ) (0 h ha\Z - t eQ a Kail the degree of merit in »umberof findtnfs (logarithmic)

° i i I 1

degrees of merit (m) given/ — - A terms. m m f o CIML/TT

Figur 5 e Graphical illustratio f quantity-qualito n y relationshipn i s

published science literature

and Ref. l/, the corresponding doubling times for the last three X level e 17.3ar s , 26.0 d 52.an , 0 years e growt.Th h e rateth f o s general analytical chemistry literature and those for the literatur f floo e w injection analysis n chromatographio , d an y activation analysi t differena s t quality level e showar sn i n tabl . Froe abov16 eth m e treatmen e mighon t t conclude that papers containing significant advances in knowledge on general analytical chemistry and/or nuclear and other instrumental analytical techniques a tinmak p yu e fractio e totath lf o n literature with a rather slower growth rate than that of the bulk of the literature. This pose serioua s s question. Doee bulf th so k publication n analyticao s l chemistry, i.e n nucleao . d ran other instrumental analytical techniques serv a scientificalle y useful purpose? Admittedly these papers represent activity,

27 Table 16

e growtTh h e literaturrateth f o s f analyticao e l chemistry, flow injection analysis, ion chromatography and activation analysis at different quality levels

Cumulativf o . no e Doubling time /td/, years Qualit, . y leve , l finding t /a f o s least/ this level »"«i*""1 Flow Injection n ChromatographIo y Activation Analysis """"""' 1936-1972 1973-1985

pl.OO Routine Q - 13 .9 1.3 1.4 2 .29 13.0 0.7S Significant Q - P 17 .3 1.7 1.8 2 .9 17.3 Important " 0 pO.50 26 .0 2 .6 2 .8 4 .4 26

Very important Q ' P0.25 52 .0 S .2 5 .6 8 .6 52

Table 17

Impact levels of papers published in the Journal of Radioanalytical Chemistry 1968-197414

Number Number of papers in the various quality groups, p of most cited papers, Year Iljapers with more than 50 citations/ 3 I - A 0 = A .75b A = 0.5C A « 0. 25d C

1968 46 17 .6 6 .8 2.6 2 1969 136 39 .8 11 .6 3.4 3 1970 255 63 .8 15 .9 3.9 4 1972 472 101 .2 21 .7 4.6 5

1974 904 164 .8 30 .0 5.4 6

at least routine, t leasa b t significant, cat least inportant, at least very important. e misdirecteb y buma tt i mucn tha i df o ht confusiod nan communication channels get clogged. Perhaps it should be thought of as noise, always present but hopefully minimized, Quality notion s "significant"a s , "important", "very important e quitar " e har o definet d a o e tried .W o t d Rescher-type categorization with the papers published in e Journath f Radioanalyticao l d Nucleaan l r Chemistre th n i y

28 1968-1984 period taking citation impact /i.e. numbers of citations a paper received in the Science Citation Index/ as a criterion of "impact" instead of "importance" Table 17 presents the results of a year-by-year categorization. As visible, the cumulative number of papers in the X = 0.25 category /at least very important/ agrees quite well with the number of highly cited ones /\ = O.234 + O.15/. Table 18 shows the bibliographical t leas"a te verth lis f yo t important papers" define d founan d d accordin procedure th o t g e outline Tabln i d. 18 e

Table 18

Most-cited articles, 1968-1932, published in the Journal of Radioanalytical Chemistry arranged in alphabetic order by first author14

No. Ration's* Bibliographic data

4 /4.96 / . 1 ADAMS , A compilatioDAMS F. , R. , f Preciselno y Determined Gamma-Transition Energief o s Radionuclides Produced by Reactor Irradiation. J. Radioanal. Chem., 3 /1969/ 99. 2. 77 /9.6/ BOWEN, H.J.M., Problems in the Elementary Analysis of Standard Biological Material. J. Radioanal. Chem., 19 /1974/ 215.

5 /9.59 / . 3 COOKSON, J.A., FERGUSON, A.T.G., PILLING, F.D., Proton Microbeams. Their Production and Use. J. Radioanal. Chem., 12 /1972/ 39.

4. 51 /3.6/ CROCKET, J.H., KEAYS, R.R., HSIEH, S., Determination of Some Precious Metals by Neutron Activation Analysis. J. Radioanal. Chem., l /1968/ 487. 5 /8.211 / . 5 GIRARDI , SABBIONI,F. , Selectiv,E. e Removaf o l Radio-Sodium From Neutron-Activated Materialy b s Retentio n Hydrateo n d Antimony Pentoxide. J. Radioanal. Chem., l /1968/ 169.

3 /6.17 / . 6 GIRARDI , PIETRAF. , , SABBIONIR. , , RadioE. , - chemical Separation y Retentiob s n Ionio n c Precipitates. Adsorption Tests on 11 Materials. J. Radioanal. Chem., 5 /197O/ 141.

-Citation/year data are in parentheses.

29 Acknowledgment

The contribution of Drs. S. Zsindely and A. Schubert and Mrs. A. Toth in the collection and svstematization of data in tables 3-14 is gratefully acknowledged.

References

Braun. T . 1 Bujdosö. ,E Schubert. ,A , Literatur Analyticaf o e l Chemistry :ScientometriA c Evaluation PressC ,CR , Florida,

USA, 1987. * Braun. T . 2 Bujdosö. ,E GritC ,CR . Revs. Anal. Chem., 13 /1982/ 223. 3. T. Braun, Fresenius Z. Anal. Chem., 323 /1986/ 1O5. 4. T. Braun, Radioisotopes, 35 /1986/ 391. 5. T. Braun, W.S. Lyon, Fresenius Z. Anal. Chem., 319 /1984/ 74. 6. M.O. Lorenz, J. Assoc. Am. Stat. Assoc., New Series No. 70 /June 1905/ 209. 7. E. Garfield, Citation Indexing, J. Wiley & Sons, 1979. 8. Science Citation Index, Institute for Scientific Information, Philadelphia, USA. 9. H. Small, Detecting Emergent Growth Areas in Science, Unpublished

report, 19-82, Institute for Scientific Information, Philadelphia, USA. . RuzickaJ . 10 , E.H. Hansen, Anal. Chim. Acta 9 /198617 , . 1 / 11. T. Braun, Anal. Proc., 19 /1982/ 352. . MoravcsikM . 12 . Res,J . Policy, 2 /1973 / 268. 13. N. Rescher, Scientific Progress, Univ. of Pittsburgh Press, Pittsburgh, 1978.

14. T. Braun, E. Bujdosö, W.S. Lyon, J. Radioanal. Nucl. Chem., 1 /19859 / 419.

30 Appendix 1 List of instrumental analytical techniques used in paoers published in the 1981-1984 period

AACT alpha activation AA5 atoaic-absorption spectrophotometry ACPD a. c. polarography ACSV a. c. stripping vol tirncri*t ry AE5 atonic-«Bission spectrometry AESP Auger electron spectronetry AFS atomic-fluorescence spectrometry AMP anperonetry APIS alpha-particle induced gamnia-ray emission ASNP anodic stripping neopolarography ASP alpha-spectrometry ASPC a I pha-spectrorn*try ASTV anodic stripping voltanmetry AUTR autoradiography BAA bicyclic activation analysis BEXE beta-excited x-ray emission CECH cation-exchange chromatography CFENI continuous flow enthalpm^try CFSS coherent forward scattering CHFA charged particle activation CLUM chewiIuainescence COND conductometry COUL cou Ionetry CPOT chronopotentiornet ry CVAM clironovol tarnmetry DACT deuteron activation DNC delayed neutron counting DPAS differential pulse anodic stripping voltawnetry DFCS differential pulse cathodic strippin tarurnetrl vo g y DPP differential pulse polarography DSCA differential scanning calorimetry DTA differential thermal analysis ELMP electron «icroprobe EMSP emission spectrometry ESR electron spi r paramagnetic(o n ) resonance FANÇ f.a.n.e.s. POMS fitld dvsorption mass spectrometry FLIM f luonrnetry FLIN flow injection analysis FSDP fast Scan differential pulse polarography FSSP forward scanning spectroruetry FTC fission tracl> counting FTSP flow through strippirig potentlorwetry GACT gariiFua activation GC gas-chromatography GEC ge I -chr ofuatography GLC gas-liquid chroniatography GPLS giant pulse laser spectrocheriiistry GFA ganirna-ray attenuation GSP gafnft.a sp*.; trorriet r y HACT heav n activatioio y n HEAC he 11urn activation HGMS hydride generation mass spectrometry HPLC high performance liquid chromatography ICHR lonchromatography ICP inductively coupled plasma ILUM indirect luminescence method INVV inverse voltammetry 10&B ion-beam oornbardenient IONF ion f I otat i c>n IONM loncni crop robe IFSP intraresonance spec trônât ry IPSP int r a r e son an ce spec t r orne t r y

31 ISDI Isotop« dilution ISE ionsetective electrode KIN kinetic methods LAPI astrr-aI t OK Uc photoionisatio n LCHR liquid Chromatograph/ LEIS laser-enhanced-ionisation spectror«etry LEPS low energy photon spectrometry LIFL laser induced fluorescence L ILL! laser induce unities'I d : en^:« LMPM laser rnicroprobe mass analysis LMSP laser niicro-spectrometry LSCH linear sweep chronoarriperometry LSPI laser stepwise photoionisation HAGN magnetometry MECS mol. «Mission cavity sptctrometry MPE magneto paper el ectrophoresis MS »ass spectroMtry HTES metastabl e-transf er emission spectrometry NAA neutron-activation analysis NCGS neutron capture gamma spectrometry NCPG neutron captured prompt gamma-ray analysis NIAU neutron induced autoradiography NISC neutron indirect scattering NMP nuclear microprobe NMR nuclear magnetic resontnct NTD nuclear track detection NTES non thermal excitation spectrometry NTIM negative thermal -ionisation mass-spectrometry OPOL osci I I opol arography PACS photoacoustic spectrometry PACT proton activation PCH paper chronatography PESP pulse excitation spectroscopy photo-fissio ayed-neutrol nde n counting PHAC photonact i vat ion PHOT photometry PIXE particle induced x-ray emission PLES plasma emission spectrometry PLU photol uwinescence POL polarography POT potent iometry PPOL pulse polarography PQCP piezoelectric quar: crystal PRAC proton activation analysis PRPS proton induced prompt-photon spectronetry PRXE proton induced x-ray emission PST potent iometric stripping method PVAM pulse voltammetry RADC radiochemistry RADM radiometry PIXE resonance ionisatiot X f no RNC radiative neutron capture RRSP resonance Ratnan spectri^metry RSSP Raman scattering spectroscopy SCHR size exclusion chrornatography SCVA stripping(chrono) vol taametry SIMS secondary-ion mass spectrorwy r t SLPI stepwise laser photo-ionisation SPCT spectroscopy SPFL spec t r of I uori met r y 5PGR spectrography SPOT stripping potent! omet r y SPPH spectrophotoiaetry SQWP square wave polarography SSCA single sweep chronoamperomty r t THCH thermochemistry THM thermometry TITR titrimetry

32 TLC thin-layer Chromatograph/ TLSF1 thermolense spectrophotc-metry TMT therraometric titrir»*try TNA thermal neutron absorption VAMM voltaiMMtry XDir x-ray powder di f fractoruetry XESP x-ray «Mission spectrometry XMP x-ray microprobe analysis XFES x-ray photoel «>:tron spectroru*y r t XRF x-ray fluorescence XSEL x+-selective elee trod« XSPC x-ray spectroscopy

33 Appendix 2

Merging of instrumental analytical techniques into groups /see appendix 1 for abbreviations/

Optical methods

AAS AES AESP AFS CFSS CLUM ELMP EMSP ESP FANE FDMS FLIM FSSP GPLS HGMS ICP ILUM IOBB IONM IPGP IPSP LAP I LEIS LEPS LIFL LILU LMPM LMSP LSPI MECS MS MTES NMR NTES NTIM PACS PESP PHOT PLES PLU RPSP PSSP SIMS SLPI SPCT SPFL SPGP SPPH TLSP XDIF XESP XMP XPES XSPC

Nuclear methods

AACT API G ASP ASPC AUTP BAA BEXE CHPA DACT DNC FTC GACT GPA GSP HACT HEAC ISDI NAA NCGS NCF'G IIAU NISC NMP NTD PACT PFDN PHAC PIXE PPAC PPPS PPXE PADC PADM PNC TNA XPF

Electrochemical methods

ACPO ACSV AMP ASNP ASTV COND COUL CFOT CVAM DPAS DPCS D F F FSDP FTSP INW ISE LSCH MPE OPOL POL POT PPOL T PS FVAM SCVA SPOT SDWP SSCA VAMM XSEL

Chromatography

CECH GC GEC GLC HPLC ICHP LCHP PCH SCHP TLC

Mi see I Ianeous

CFEN DSCA DTA FLIN IONF KIN MAGN F OCR PIXE THCH THM TITP TMT

34 Appendi3 x Country groups

US CanadA& a Eastern Europe

Canada Bulgar la USA Czechosloval la German Dern.F'ep. Hungary United Kingdom Pod Ian F'oriian i a United kingdom Yugosl avi a

Fed.Rep.Germany North Afric Nea& a r East

Fed.Rep.Germany AIgerla Egypt Iran France Iraq Jordan France Li ban on Saudi Arabia Tunisia USSR U.Arab Emirates

USSR Rest of Africa

Japan Ethiopia t enya Japan hozzarubique Nigeria Sierra Leone Asia

Hong [ ong Rest of Developed Countries India P.R.China Austraa l l Patistan Ausia rl Singapore Be l g i urn South Korea Denmark Taiwan Fini and Greece d an I e Ir Latin America Israel Italy Argentina Nether land Brazii New Zealand Chile Norway Cuba Portugal Mexico South Africa Venez Lie I a Spain Swed en SwitzerI and y e T l u r

35 DISCUSSION

C.J. Pickford. Why was lead omitted from the analysis: wasn't the division f techniqueo a littls e arbitrary?

. BraunT e investigatioTh . s startewa n d with abou 8 elements2 t A . preliminary check-up indicated that the ranking doesn't change in case we reduce the list to 15 elements. The division of techniques we did base our e devise e Analyticath ranking th d uses y an dwa b d n o sl Abstracts database.

W.D. Shults worthwhile b . t Woulno t i o studd t e demographe th y f o y instrumentation purchases in order to measure use of different methods in additio o publicationt n s (which reflect developmen f differeno t t methods)?

. BraunT .facn i Thi ts donewa s . y talkm Som f o es (not orae showth ln i n presentation but included in the printed text) are collecting and displaying such type f datao s .

J. Tölgyessy. What is your opinion about the usefulness of nuclear analytical methods for the determination of organic species?

T. Braun. The recent study is dedicated exclusively to elemental analysis. d usefulnesan e us f nucleae o sTh r analytical techniques (compare o othet d r instrumental techniques s bee )ha r previou nou dealf o e st on witstudie n i h s published in Radioisotopes (Japan) in July 1986.

Hu attemp. yo Ross d o breaDi t . k down growt d usaga functio an hs a e f o n speed, cost, sensitivity, accurac d multielemenan y t analytical needs?

. BraunT . No .

36 USE OF SCffiNTOMETRICS TO ASSESS NUCLEAR AND OTHER ANALYTICAL METHODS*

W.S. LYON Ridgk Oa e National Laboratory, Oak Ridge, Tennessee, United States of America

Abstract

Sei entome trie sf quantitativ o involvee us e th s e methodo t s investigate science viewen informatioa s a d n process. Sciento- metric studiee usefub n n ascertainini lca s g which methods have been most employed for various analytical determinations as well as for predicting which methods will continue to be used in the immediate future and which appear to be losing favor with the analytical community. Published papers in the technical litera- e primarth ture yar e source material r scientometrifo s c studies; statistical methods and computer techniques are the tools. Recent studies have included growth and trends in prompt nuclear analysis impact of research published in a technical journal, and institutional and national representation, speakers and topics at several IAEA conferences, at modern trends in activation analysis conferences t othea d r,an non-nuclear oriented conferences. Attempts have also been mad o predict e t future growt f variouo h s topic d techniquesan s .

INTRODUCTION How does one go about choosing an analytical method for a particular determination e naivTh ? e answe s thai re selecth t s the best. Actually, bes r manfo ty analyst e methoth s di s that they have available. Activation analysts advocate activation analysis; mass spectrometrists swea y masb r s spectrometry; atomic absorption people hold hige flamth h e o s aloft d An . forth. There have been, of course, comparisons of analytical techniques for different elements in certain matrices. TABLE 1(1) shows typical dat sucn i a a comparisonh e numberth ; e ar s usually calculated on a best case scenario. Such tables are most useful for demonstration or comparisons. No one can really take them seriously except where differences are clearly large, or one method has an order of magnitude advantage. Different matrices require different methods. TABLES II & 111(2) show

* Research sponsored by the Office of Energy Research, US Department of Energy, under contract DE-AC05-84OR21400 with Martin Marietta Energy Systems, Inc.

37 Table 1 ELEMENTAL SENSITIVITIES BY SEVERAL TECHNIQUES i Element Sensitivity (nanogtams/gcam)

1 NAA1 AA2 ICP-AES3 SSMS

*9 0.1 2 0.7 0.2 Al 1 20 0.6 0.02 AS 0.5 20 3 0.06 Au 0.05 10 - 0.02 B - 2,000 1 0.01 Ba 10 10 0.1 0.2 Be - 2 0.01 0.008 Bl - 3 8 0.2 Ca 500 1 1.6 0.03 Cd 20 2 0.1 0.03 Ce 10 - - 0.1 Cl 0.1 - - 0.4 Co 0.9 10 2 0.05 Cr 200 3 0.4 0.05 Cu 0.2 1 0.06 0.06 F 0.1 - - 0.05 Fe 3 1Û x 0 4 10 0.5 0.05 Ga 0.4 100 5 0.09 Ge 2 200 10 0.2 Hg 2 200 20 0.06 K 40 1 1.0 0.03 Mg 50 1 1.5 0.03 »In 0.005 2 0.02 0.05 No 30 600 2 0.3 Na 0.9 0.2 7 0.02 Nb 100 200 9 0.8 Nl 50 2 U. S 0.7 P - 3 10 < 0 10 8 0.03 Pb - lü 2 0.3 Pt 5 100 - 0.5 <; 20 x 103 - 0.2 0.03 Sb 10 40 4 0.2 Sc 50 20 0.4 0.04 Se 50 100 2 0.1 Si 20 x 103 20 1 0.03 Sn 20 20 3 0.3 Sr 0.9 2 0.04 0.09 Te 9 50 4 0.3 Th 10 - 8 0.2 Tl 10 50 0.6 0.5 Tl - 20 4 0.2 U 0.4 3 10 3x 0 25 0.2 V 0.1 50 0.6 0.04 W 0.7 1 x 103 3 - Ï 1 x 103 100 0.1 - Zn 20 1 0.2 0.1 Zr 200 1 x 103 0.7 0.1 (1) G.H. Mo r r i son in Elemental Analysis of Biological Materials, IAEA Technical Reports Series 2 (I960)19722 . ,p . . f Kitkbcight . (2G ) , Ibid . p 155, . . . J FasselHaas A . (3. W )V , , Ibid . p 170, .

38 TABL . II ECompariso f INAo n A with IDSSM r determinatiofo S n of Ba, Cr, and Sr, and with He Arc emission for determination of As

Significance Procedure N ug/ ga gg/g vg/g t Probabilities*

Ba INAA 42 155.6 42.3 94 232 3.19 0.01 IDSSMS 42 132.3 46.0 40 220

Cr INAA 45 29.54 6.6 21 59 2.61 0.011 IDSSMS 45 24.918 4 10. 7 6

Sr INAA 43 115.7 31.7 46 185 0.097 0.923 IDSSMS 45 114.9 37.6 44 230

As INAA 31 29.66 11.8 12.4 58 5.61 0.001 He Arc 31 36.24 13.3 16.2 65 •Probability that the t value is due to chance alone. Values <.01 are taken as indicating a significant difference in the calculated" means of the analytical results for a given element.

TABLE III. Comparisod IDSSMan r A determinatiofo A S f o n n of Cu, Zn, and Pb in coal

X 'rain Jjnax Significance Procedure N Vg/g o pg/g vg/g t Probabilities*

Cu AA 23 21.77 1.73 17.3 25. 0 1.18 0.25 IDSSMS 3 2 20.22 4 6.0 32 1

Zn AA 23 24.90 7.87 14.2 40 0.34 0.74 IDSSMS 23 24.09 8.46 10 40

Pb AA 23 13.22 4.67 8.2 30 1.33 0.19 IDSSMS 23 11.43 4.44 5.0 21 »(See r explanationTablfo I V e )

39 some actual experimental comparisons performed at ORNL for the determinatio f severao n l element n coa i sy severa b l l methods. Data such as these can be useful in deciding what method to use for specific problems, but time and effort is required to make these experiments. From a philosophical or theoretical point of view it would e nic b o knot e w what method e beinar s g most user analysisfo d , what trend e occurringar s , whethe w techniqune a r catchins i e g on or whether an old technique is being replaced. Researchers would lik o knot e w whic w methodne h e deemear s d most important by their peers; they would e 'wherb lik e o actiot eth e n is'. Publishing papers on a subject in which no one is interested is more likel o leat yo oblivio t d n tha o advancementt n . Seientometries is a term representing those quantitative methods that deal with science viewed as an information process. Scientometric s beeha sn use o studt d y information processea n i s multitud f disciplineo e s including physics, chemistry, analytical chemistry, biology, health physics and astronomy. Informa- tion flo n patentsi w n conferencesi , n journali d s ,an bee ha s n elucidated through scientometric techniques. Primary sources r thesfo e e publicatiostudieth e ar s n records from technical journals and books; statistical and mathematical techniques are e toolsth n excellenA . t general referenc e methoth d o an dt e many of its applications is the monograph by Braun and Bujdoso (D- e res Th f thio t s paper wil e concerneb l d with giving some example f scientometrio s c studies that e problerelatth o f t eo m assessing different analytical methods. The thrust of the s possibli t i effor o shoo t predicw t e ho s wi t t which methods are more likely to be successful and in what direction analytical chemistr e goingb y ma .y Obviously scientometric methods can never take the place of experiment for a particular application e useb n polict thei n dbu , ca y y n makini d an g research planning, evaluatio d predictionan n .

P ADVANCETO ANALYTICAN I S L CHEMISTRY, 1935-1985 A primary too n scientometrici l e SCIENCth s i s E CITATION INDEX (SCI) publishe y ISI b n thesdI . e volume e collectear s d all the citations to articles in past years in all the major technical journals. Thus, if one published an article in ANALYTICAL CHEMISTRY in 1984, and I referenced it in an article f mino n 1985i e , this citation would appea o yout r r articln i e s apparen i e SCI t th I . t tha e morth t e heavily cite n articlea d , e morth e impace fields havini th t i n to . g e werThuson o t ef i , select the papers that had the greatest number of citations over a perio f timeo d e would,h n theoryi , , have selecte e mosth dt important advances in the field. Of course, another way to ascertain top advances would be to poll a number of influential experte fiel th d ascertai n an di s n what they considered were th e best achievements.

40 d whajuss di Brau wa ) t (4 ndescribe d above e useI h SC :d to find the most cited papers in analytical chemistry while at e samth e tim sene h e circulaa t r analytica6 lette18 o t r l chemistry gatekeepers exerto gatekeepeA ( wh . e s on influenc s i r e on a field as editor, reviewer, etc.) TABLE IV is a merged summary of the top advances as seen by the analytical chemists. Fro mlisa f moso t t cited I BraupaperSC n n i deduces e th d methods used in these most cited papers; TABLE V gives his results. Ther s i considerable e agreemen t seto betweetw s e th n of data which seems to lend support to the use of citation data for evaluating progress in a field. It is noteworthy that nuclear analysis ranks fourth among the top five in both compilations. That in itself should tell us something about the value of these techniques relative to others.

TABLE IV. Merged summary of top advances as seen by analytical chemists

Rank Topic Responses* (No.)

1. Advance spectroscopn si spectrometrd yan y 156 2. Chromatography 127 3. Advance electroanalytican si l chemistry 79 4. Nuclear analytical chemistry 77 5. Continuous flow analysis 23 6. Surface analytical chemistry 22 7. Immunoanalytical chemistry 21 8. Chemometrics 18 9. Polydentate complexing ligands 14 10. Organic reagents 6 Thermoanalytical chemistry 6

e.g. different advances in spectroscopy and spectrometry were mentioned a total of 156 times in the responses of the 69 re- spondents

TABLE V. Methods used in 133 "most-cited" analytical chemistry papers f totao l% Rank Topic

1. Spectroscop spectrometrd yan y 44.8 2. Chromatography 36.8 3. Immunoanalysis 13.8 4. Nuclear analysis 3.4 5. Electroanalysis 1.2

41 IMPAC F RESEARCO T H PUBLISHE N JOURNAI D F RADIOANALYTICAO L L CHEMISTRY. w tur no o somt ns u e t specifiLe c radiochemical researcd an h see what the top research papers in this field have been. Braun, Bujdoso and Lyon (5) used all the research papers publishe e JOURNAth n i F dRADIOANALYTICA O L L CHEMISTRY (JRACn )i the years 1968-198 e datth a s a 1base . Total numbe f citationo r s o thest e papers were obtained usin e SCIENCth g E CITATION INDEX. Tablx mos si I list V ete citeth s d paper n alphabeticai s l order by first author. This list of papers seems to provide a neat summare mosth tf o yactiv e area f researco s n nucleai h r analytical chemistry. Number one is the use of hydrated antimony pentoxide for removal of sodium in activation analysis. When this method was first reported, the authors (6) of the bi-yearly Nucleonics articl ANALYTICAn i e L CHEMISTRY wrote, "The most exciting find in chemical separations, however, was the discovery that the radio-sodium interference, which has haunted biological analysis since the beginning of neutron activation, can be removed by a column of special grade hydrated antimony pentoxide." The other papers cover such topics of interest as radiochemical separations, gamma ray energies, proton microbeams, standard reference material d preciouan s s metals. Probably s woulu mos f o td agree that these were very important topics at the time, and for that matter are still important.

TABL . MosVI E t cited articles 1968-1982 published in JRAC

1968-1982 No citations' Bibliographic data . 1 64(49) ADAMS , DAMS F. Compilatio, A , R. , f Preciselno y Determined Gamma- Transition Energies of Radionucbdes Produced by Reactor Irradiation / RadioanaL Chem (1969)3 , 99

2. 77 (9.6) BOWEN, H. J. M , Problems m the Elementary Analysis of Standaid Biological Material. / Radioanal Chem, 19 (1974) 215

) J 9 ( COOKSON 5 9 FERGUSON. A 3. PILLING, .J , G . . ProtoT D. . A , F ,n Mictobeuns. Their Production and Use / Rfdioanal. Chem (W2)12 . 39

4. 51(3.6) CROCKET, J H., KEAYS, R. R , HSDEH.S., Determination of Some Precious Metals by Neutron Activation Analysis / RadioanaL Chem., (1968}1 487

S 115(8.2) GIRARDI, F., SABBIONI, E .Selective Removal of Radio-Sodium From Neutron-Activated Material Retentioy !b Hydraten no d Antimony Penloxide J. Raaioanal. Chem., (1968)1 169

6. 73(6.1) GIRARDI, F., PLETRA. R, SABBIONI, E., Radiochemical Separations by Retention on Ionic Précipitât«. Adsorption Tests on 11 Materials. J. RadioanaL Chem , 5 (1970) 141

•Citations/year data are in. parentheses

42 F CONFERENCO E US E PAPER O DELINEATT S E TRENDS Papers presented at conferences and national society meetings represent the cutting edge of science. Here is where advance e firsar s t reporte o peerst d e ModerTh . n Trendn i s Activation Analysis Conferences have been a primary vehicle for reportin n thai g t field since 1961. Lyo) looke(7 n t papera d s and subjects included in five of these conferences between 1961 and 1976. He also attempted to predict the product mix of e 198paperth 2t conferenca s e (8) e latte.Th r effor s onlwa ty marginally successful. The data analysis of papers at the 5 conference d shodi sw some interesting trend s seea sn FIGUR i n E e rapiTh d. 1 risd fal f an eintereso l n neutroi t n generators i s easil ya simila s seeni s ,a r ris f chargeo e d particle techniques. This paper also pointed out which countries and laboratorie a growin d ha sr declinino g g interes n activatioi t n analysis, and suggested material science and industrial applications as the areas of most interest over the years. FIGURE 2 shows the percentage papers by category and is worth studying as a sort of historical record of scientific interest during the period.

FOREIGN NORTH AMERICAN

40 0 40 NEUTRON GENERATOR 0

UJ Occ 40 UJ CHARGED PARTICLE Û. 0 40

0 PHOTON 40

0 SOURCE

61 65 68 72 76 YEAR

Figure 1. Paper Categorization by Technique

43 FOREIGN NORTH AMERICAN

MATERIAL SCIENCE, INDUSTRIAL

PLENARY

ACCURACY, SPEC, STANDARDS

ENVIRON. AND ECOLOGY

BIOLOGY- BIOMEDICINE

GEOLOGY, BIOCHEM.

SEPARATIONS

6 7 2 7 8 6 5 6 1 6 YEAR

Figure 2. Paper Categorization by Subject

Other studies concerned with conference proceedings have been made by this author. The IAEA Nuclear Techniques in the Life Sciences conferences were studie d comparean d o othet d r similiar meetings (9) e ultimat.Th e fat f papero e s presentet a d a national American Chemical Society meeting was ascertained through a literature survey (10). About half the papers presented orally were eventually published somewhere.

PROMPT NUCLEAR ANALYSIS: GROWTH AND TRENDS Prompt nuclear analysi s definei s a revie n i d w paped en r two bibliographies as the use of prompt radiation accompanying a nuclear reaction to measure elemental or isotopic concentrations (11). Data from these were a studuse n o i assesdt y s growtd an h trends in the field by Bujdoso, Lyon and Noszlopi (12). Various methods were evaluated and a great wealth of data were generated using citation data from SCI. FIGUR 3 showE e growtth s f o h publications in this field over time, while FIGURE 4 shows distribution of papers on various types of nuclear reactions. From thes o figuretw e e see on n sincreasina s g interes n prompi t t nethod s wel a ss whic a l h nuclear reactions wer n favoi e r over the years. Backscattering seems ever popular. Oxygen, silicon,

44 fc 10' W 9 S î 'S

• • • • • •

1 I I I I J_ 1952 1956 1960 196t 1968 1972 1976 1980 Years

Figure 3. Publication Output in Prompt Nuclear Analysis

1948 I960

Figure 4. Distribution of Publications on Various Types f Nucleao r Reactions

45 carbon, aluminum and nitrogen make up the first five most popular elements determined by prompt methods. Backscattering is larger than any other single charged particle technique, but makes up only about 1/3 to 2/3 of the total per element. TABLE VII show e percenth s t makeu f papero p n analyticai s l chemistry and prompt nuclear analysi y countryb s . Three countries account for over 2/3 of the prompt papers. Little correlation is seen between the two groups. Careful study of the data in this paper will reward the researcher who may be thinking of entering the field r examplfo ; e after three yearl f researchal o s f o % 68 , publishing workers had left prompt nuclear analysis!

TABLE VII. Percent makeu y countrb p f papero y n o s prompt nuclear analysis and analytical chemistry

Percentag f papere» n so Country prompt nuclear total analytical analysis chemistry*

USA 41.7 18.0 France 13 3.0 United Kingdom 10 5.4 Fed. Rep. Germany 4.8 5.7** Soviet Union 4.6 25.5 Canada 3.5 Sweden 2.9 South Africa 2.8 Belgium 2.4 Australia 2.3 Japan 2.1 8.8 Czechoslovakia 1.6 4.9 Italy 1.4 1.7 Poland 1.3 2.4 Denmark 1.2 Finland 0.6 Hungary 0.6 Netherlands 0.5 1.0 India 0.4 3.7 Venezuela 0.3 New Zealand 0.3 Rest of the world 2.0 19.9

•Average of the years 1960-1970.' * ••Includes both EasWesd tan t Germany.

NUCLEA D OTHEAN R R TECHNIQUE R POLLUTANFO S T ANALYSIS Most of the work that this author has done has been concerned primarily with nuclear techniques, consequently the s centerediscussioha s r nowu n fa themo t d , o Le s n. however , a recenloo t a k t comparison performe y Braub d n (12) e use.H d articles in ANALYTICAL ABSTRACTS to ascertain techniques used

46 r pollutanfo t analysis e founH . d that through 1984 chromatography accounted for 33% of papers published, spectrometry 31%, electroanalysis 19% nuclead ,an r 16%. These percentages agree rather well wite datth ha show n TABLi n . V E Pursuing the comparison furthur, Braun tabulated the instrumental methods used in an IAEA (64 lab) round robin intercompariso f elemento n n environmentai s l water samplese Th . result showed atomic absorption leading (53%). Neutron activation analysis was a distant second (22%), with other optical spectrometries (15%) and mass spectrometry (5%) filling e remainin th t mos ou f o t g techniques. Still another survey type approach to this question is to e whicse h instrument e mosar s t availabl n differeni e t laboratories. This almost certainly removes nuclear techniques from the race. Allied to this type survey is that often published in trade magazines wherein labe quizze o whaar st s t a d instrumenta- tioe cominth n n thei g y year bu pla o .t n Durin e pasth g t decada half d an ,e increasing emphasis ha s been place n elementao d l speciation. Here nuclear techniques are at a severe disadvantage because any speciation done must be by a chemical separation. Braun has shown that between 1975 and 1985 publications involving speciation increased about threefold. Finall t shouli y e pointeb d t thae presenou d th t t trenn i d analytical instrumentation is toward sensors that can be placed directly in the environment and which make real time continuing measurements n linI . e gamm y measurementra a s havr e fo gon n o e decades, but more sensitive and selective techniques are under development.

FUTURE AREA R INVESTIGATIOFO S N The purpose of this paper has been to introduce the subject of scientometric study to the reader and to indicate how it has been use o evaluatt d e progres n sciencei s e fielTh f .o d analytical chemistr s idealli y y suite r sucfo d h studies: there is an excellent literature composed of only a few top journals in which researchers publish, data on samples and methods are available from laboratorie d governmenan s t agenciesw ne d ,an method d instrumentan s e constantlar s y beinge addeth o t d chemist's repertoire. On-line computer searches are easily and inexpensively performed. Results are often quite surprising and informafive. For those old-fashioned enough to want to work in the laboratory, the results of such studies will provide insight and direction as to where methods are needed and where the field is f sensoro e headeds beeus ha s ne Th mentione. a fruitfu s a d l area for study. Robotics is closely allied to sensors and is certainly an area of future growth. The analytical radiochemist is no longer the "new kid on the block;" there are many newer and often cheape d bettean r r method o perfort s m elemental analyses. The challenge to the radiochemist is to find where his technique besbe the and tn can usesimprov ... d e them.

47 REFERENCES ) LYON(1 . S.,,W " Practical Application f Activatioo s n Analysis and Other Nuclear Techniques," Proceedinge th f o s International Symposiu n Applicationo m d Technologan s f o y Ionizing Radiations, Riyadh, Saudi Arabia, 12-17 March 1982. King Saud University Libraries (1983) 871. (2) LYON, W. S. et al, "Analytical Determination and Statistical Relationships of Forty-one Elements in Coal from Three Coal-Fired Steaoi Plants," Nuclear Activation Technique e Lifth en i Sciences s 1978, IAEA, Vienna (1978) 615.

. ) Revt BRAUN(3 i .Cr , BUJDOSO C T. ,Anal CR , .) ,E. Chem (3 3 1 . (1982). (4) BRAUN, T., Fresenius Z. Anal.Chem. 323 (1986) 105. (5) BRAUN, T., BUJDOSO, E., LYON, W. S., O. Radioanal. Nucl. Chem. 91/92 (1985) 419. ) , RICCILYON (6 S. , Anal.Chem , ROSS. H. W ,E. , . ,H (19702 _4 . ) 123R. Radioanal. J ) , LYON(7 S. . ,W . Chem (19829 .J5 ) 107. ) , RadiochemLYON(8 S. . W , . Radioanal .Lettr (19822 5_ . ) 369. (9) LYON, W. S., Sei en tometri es _2 (1980) 215. (10) , ROBERTSLYON S. , Anal . P. W , . .P , Proc. (London (19830 )_2 ) 374. (11) , CAMPBELLBIRD R. , CAWLEY , . PrompL. ,J J. . . B ,R ,t Nuclear Analysis Bibliography, Australian AECRE Lucas Heights AAECI 3 (1978)44 E . (12) BUJDOSO, E., LYON, W. S., NOSZLOPI, I., J. Radioanal. Chem. 1± (1982) 197. (13) BRAUN, T., "Charting the Trends in Nuclear Techniques for Analysi f Inorganio s c Environmental Pollutants," IAEA Expert Meeting on Nuclear Methods for Analysis of Inorganic Environmental Pollutants, Karlsruhe G (1986FR , pressn )i .

48 COMMENT ESTABLISHMENE TH N SO A F TO METHODOLOGY FOR THE METROLOGICAL CHARACTERIZATION OF ANALYTICAL METHODS

E.H. KLEHR Oklahoma University, Norman, Oklahoma, United State f Americso a . TÖLGYESSJ Y Slovak Technical Univers ;ty, Bratislava, Czechoslovakia

Abstract

e comparisoth r Fo f analyticao n l methods s importani t i , t a standardize e tus o f statisticao t se d l techniquee th r fo s evaluation of data and for the comparison of methods. There seems to be some confusion regarding the definitions and calculation r basifo s c terms suc s accuracya h , precision, detection limits, etc. Currently, these concept t havno eo d s commonly accepted definitions or methods of calculation. It is proposed that we work toward the establishment of a common method of presenting such data, and a common method of its calculation. Only in this way will we be able to compare various analytical methods in a meaningful way. In this paper a ,proposa a unifie s madr i l fo e d methed- ology for the metrological characterization of analytical methods.

. I Introducn tio maie th n f o characteristic e On f traco s e analysis i s the need to identify and determine a large number of elements and chemical species, many at very low concentrations, and in sample f greatlo s y differing composition. Because th f o e variet f analyteo y d samplan s e types a ,numbe f differeno r t technique e neededar s , including nuclear analytical methods. A qualified analyst must be able to choose which method is best suited for a particular problem, and to do so effectively a commonl, f figureyo uset f se merido s t shoul e availb d - able. Unfortunately f figureo t sufficientl ,no t sucs se i sa h y developed at present. This represents a need which merits serious attention, especially for those interested in nuclear analytical methods.

As examples of figures of merit, we suggest considering concepts suc: samplas hd samplan e e pretreatment requirements, total cost, total analysis time, accuracy, precisione th , abilit a metho f o yo detec t d t and/or quantify small amounts or concentrations, and sensitivity.

Som f theso ee difficultar e r eveo , n impossibl o quantifyt e , such as sampling requirements. Others may be amenable to standardized evaluation processes, suc s accuracya h , precision,

49 e abilitth d o detect/quantift an y y small amountw conclo -r o s entrations of analyte in samples of varying composition. Of these e havw , e chose e lasr discussionth nfo t .

e IUPAC/ACTh 2. S approach The International Union of Pure and Applied Chemistry has provided the following definition [l]: "The limit of detection (LOD) , expressed as a concentration or amount, is derived from the smallest measure that can be detected with reasonable certainty for a given analytical procedure." The American Chemical Society [2] gave a similar definition: "The limit of detection is the lowest concentration of an analyte tha n analyticaa t l procedur n reliablca e y detect." Before such concept e usefular s , terms suc s "reasonabala h e certainty" and "reliably" must be quantified.

The IUPAC and ACS have used similar approaches in quantification ,e presente b whic n ca h s followsa d e limiTh f :o t detectio a numbe s i n r that describe e lowesth s t concentration or amount of analyte that can be determined as statistically different fro n analyticaa m l blank wit a certaih n degref o e confidence n actuaA . l measuremen s usualli t y mad n termi e s oa signaf l (volts, numbe f countso r , area unde a spectrumr ) which mus e relateb t a concentratio o t d r amouno n f analyto t e through some response function, a calibratiosuc s a h n curve. Thus we can write: S = S + S or S = S - S where = total S „ r 1sampl o , A,e signalD , A. I ,ü o analyte= signat t signal e S ne du lr o , , S = signal due to blank, or background. The response function can be written as: S = g(C ), a concentratio s wheri C e r amouno n f analyteo t g rep d -an , resents some function. In the simplest linear case, we have:

Sx = aCx; and C = (ST - Sß)/a. s assumeIi t S wil, e randomldb S l tha S t y distributed according to some well-defined distribution function (eg. Normal, Poisson, Student-t, etc.).

Hence our ability to determine small amounts of chemical species depends on distinguishing small differences between t thi n Bu turi s . n S depend d an n fluctuationo s S S n i s S values e morth ; e they fluctuat r repetitivfo e e measurementa sample f o so distinguise lest th ,e sw able ar e h them. A common measur f fluctuatioo e e standarth s i n d deviation, defined aj^L.____ s =^JTS - S") /(n - 1), where e arithmetith s 2i " c averag n replicat f o e e measurementsf I . the random error S follo n i s a normaw l distribution a plo, t of frequency of observations vs signal size would appear as in Figur . Obviously1 e e abilitth , o distinguist y h between analyt d blanan e k depende differenceth n o s- S~) r S .( :Fo instance, in the case of a normal distribution, if S is 3s larger than S— (average of blank signals), there is a 0.13% probability that the signal is from a blank. This small probability is taken to fulfill the requirement of "reasonable certainty r "reliably"o " . More generally n writeca e w ,: a norma r fo l 3 distribution = k ; s k .+ — S = S D B DBU

50 FIG.l. NORMAL DIST. OF ANAL. SIGNALS

Kaiser [3] has proposed a more conservative value: k = 6, and propose e termth d : limi f guaranteeo tS wen n AC o te Th . to define the limit of quantification as follows: S = Ss g+ k_s,,-- + k„s„k ; ; k„ = 10 for a normal distribution. D j t D e Iassumw f a simple e linear response function, these o signatw l limits translate into concentration amount: as s CD ' kDSB/a CQ - V3' A number of variations of this basic approach have been suggested [4], including: f Student-o e r caseUs fo 1. n plac k i st f whero ee th e numbe f replicato r e measurement s smalli s t shoulI . e b d noted that unlike the normal distribution, t values depend on the number of observations. It should be chosen to give the same confidence level as k (eg. 0.13% for LOD). 2. Use of Student-t and the standard deviation of the average blank signal: s— , = numbe s /ifnin f = blan o r : - s Dk measurementsts D .B In many cases, 3s > ts-(o^= 0.05), and hence the use of s- a significan y b C n reducr ca o t amountC e . Thereforet i , is important to state clearly which approach is being used. e pooleth f o d e standarUs . 3 d deviation, define: as d

In this case, it is convenient to assume s = s„, but this must be verified experimentally. If the assumption is valid, y writema e w : \

e relativth f o e Us standard deviation, RSD, defined as RSD SB/S£, and SD "Vu = I/RSD k . e signan thiTh i s S lcas ee so-calle leadth o t s d concentration limit oT detection. This RSD variation is related to the others, as can be seen from the equation for a simple linear response function: = ksB/a k(sn/S-)(S-/a )k(RSD)(S-/a)= . D D D D

5l Usin e approacth g h discusse s recommendeha d S aboveAC e th ,d the following terminology: 3 ND; analyte not detected 10s B' Regio f detectiono n ; signals measured d reportean s amounta d r concentrationo s s D giveLO n witparenthesie i n th h s S.. > IDs Region of quantitation. 3. Approach using the error in sensitivity (Graphical approach) Winefordne d otheran d Lon] an sr[A ghav e pointet ou d that the IUPAC/ACS models consider only the errors in blank measurements, whereas most practical cases are more complex. Many analytical methods require the use of a response function (calibration curve), which for the linear case will now be written : sx = mCx + i- This equation is best obtained by regression analysis of a good set of calibration data. The slope, m, is a measure of e sensie interceptth th s e i tmethod ivitth i s f a d ,o y an , shown in Figure 2. Apparently some workers assume m to be well defined, but in general it mayhave an uncertainty due to e calibratioerrorth n i s n points r non-linearityo , . This e representeb erro y ma r : as d m +_ ts , e standarth s wheri ds e deviatio e slopeth f o .n Student-t is choseTun to give the desired confidence level for a given number of observations (the number of points used to define the calibration curve). Figure 2 illustrates the effect of including error m whe n i ns evaluatin e LODth g . Three concentration e a givee valuefoun b Th r n fo nd. ca s valu S f o e conservativ whic, e smalles C s baseth i h s n i o d D t LO esen -

. ) si s t ivit-t m ( y m AnalagouQ s LO remark e madr b fo n e ca s valuei s .

- — S.—

FIG.2. CALIBRATION CURVE

52 A. Propagation or erros approach A closer loo t Figura k 2 eindicate s that perhaps errors in the intercept, i, should be included in LOD, LOQ values, and to do so, it is convenient to use propagation of error theory, which leads to the equations: - i)/ mS ( = C A _ * c S K ^ '"*"* This equat ion allows the évaluât ion of the standard déviât ion of a concentrât ion/amount taken from a calibration curve.

Fo rD calculationsLO e blanth s ki signa S , l and

-•_ —= «_>_s . lU.LU-ll\_l,Further mosn i ,j tAl l caselll\~l tj s1~ V- C^l iL-J \_ O , Oî^ ~~ \J . Substituting these equalities inte abovth o e equation gives: * 2 , . , ,, +s . ( i/m) J /m i

, k= Vc anQ ,sQ - Dcd £ . If there is no error in the slope or intercept, e equatioth d an n , 0 abov = e. s simplifie = e basis th o ct s iffpAC/AcS model.

. 5 Approach usin e relativth g e standard deviation More recently, Mocak and Beinrohr [5] have presented a method for calculation of terms similar to LOD, LOQ, which uses the appropriate relative standard deviations, RSD, and knowledge of their concentration/amount dependence. They begin by defining three different limits, as shown in Figure

fcto

FIG.3. SIGNAL LIMITS

1. e signa imiTh Deci1 tn . ls io correspont e uppeth o rt s ( 1-0J confidence interval for the distribution of blank signals The Type I error is equal to ck\ ie. a signal is rejected as a blank, but it is really a blank at &J, probability. For this y writecasema e w ,: 'S B C-SB SC-SB = [S SC SB y writema e w : , s = Assumins = s g oU SB s[(n. n. SC-SB

53 If we now assume a simple linear response function e havew , ) : aC = S (

= l/a(Sc- S)

l/a(S SC-SB RSD, CC/CC

. = SSC-SB/C RSD„(%) = 100/tC. The one-sided Student-t depends on the degrees of freedom (n + n - 2) and on the confidence level, Eg. V= 6; c\= 0.05; t = 1.94; RSD = 100/1.94 = 51.1% ; (J^=6 = 0.01 = = 3.14100/3.1; t D RS ; = 431.8% .

2. Detection limit. The signal, S is such that the (lower end) confidence interva e grouS observa th f o pr fo l - tions corresponds to the decision limit, Commonly, The Type II error is equal to/3; ie . a signal is accepte a blan s a dk eve t fronno a blanthougms i t t a ki h /0% probability. A derivation similar to that above leads to the expression: RSD (%) = 100/2tD.

Eg. V= 6:

3 .Déterminâ______t ion limit . The signal S is such that the e confidencth f o lowe d en er interva f observâo l tS ion f o s corresponds to the detection limit, S . Commonly, ^= 0^.-/3. According to the authors, a signal greater than S "occurs e regioith n f signalo n s detected with certainty A deriva. - tion similar to that eiven for the decision limit gives: = 100/3t^ ) RS(% D . Eg.qV"= 6; Ct\= 0.05; RSD = 100/3X1.94 = 17.0%.

The authors used these limits to define concentration/ amount region s showa s n Figuri n . 4 e

RSCC

2.RSG

RSC

FIG.4. CONCENTRATION REGIONS

54 1. Analyt t detectedno e . Regio f concentratioo n n below e detectioth n limit. 2. Rang f detectiono e . Presenc e analyte b th n f o eca e detected with a given confidence level. The lower border e detectioith s n limit. 3. Analytical range. Results can be quantified with a given confidence level e loweTh . r e deterbordeth s i r- mination e uppelimitth rd e saman borde,th e s ha r A. Working range. Range includes concentrations RSD values less than a chosen value; eg. 2XRSD . . nun To implement the method, one must determine the RSD of signal values ove a rangr f concentrations/amountso e . Once thi. values done i D se RSD„ RS th , s d ,an RSD„ D RS , n i m j -Q D j j/-C , - can t be used to define the ranges mentioned.

e developmenTh s presentea , t d here, depend a simpl n o s e linear response function A mor. e complex function would lead to other expressions for critical RSD values. As opposed to the IUPAC/ACS methods, the various limits depend strongly on the numbe f blano rd samplan k e measurements. Thi s becausi s e the RSD approach uses the Student-t rather than the normal distr . ibu n tio

. 6 Comments The basic IUPAC/ACS approach is simple and relatively easy to implement. However, it depends only on blank values, and refers to a single sample measurement. In good laboratory practice, replicate sample measurements shoul e madeb d d an , average signal values should be used. This would, in turn, lead to lower LOD, LOQ values than those predicted by the basic IUPAC/ACS approach.

In the modification with the pooled standard deviation, accoun s takei t f replicato n e sample measurementse th t bu , assumptio t trues no mus = s e verified,i b t t s ni f I . the modification will be more complex.

The sensitivity error approach is valid only when the intercep s weli t l defined, whic s probable i generah th t no ly case .

e propagatioTh f erroo n r approac a powerfuh e seemb o t sl method t sufficienbu , t replicates mus e takeb t o thae s n th t normal distributio s validi ne Student-Th . t distribution e morb ey ma appropriate.

The approach of Mocak and Beinrohr requires the greatest amoun t seemf i datao t t s bu , capabl f providino e e greatesth g t amoun f informationo t . Further studie e needear s o assest d s the utility of their method.

It is not possible at present to give a definitive compariso f theso n e approaches s certainli t i d an y, premature to maky decisioan e o whicht e s besf a anynr th i , o t s i , mos te compariso usefuth r fo l f nucleao n r analytical techniques. Nevertheless, the ability to detect/quantify small amountw concentrationlo r o s f analyto sf crucia o s i e l impor-

55 tanc n environmentai e l studies A wel. l defined, concensus figure of merit for this concept would be of great benefit in discusin e applicabilitth g f nucleao y r techniqueo t s environmental analysis, and help greatly to establish a rational basis for the comparison of nuclear analytical methods with competitive techniques.

REFERENCES

[I] Spectroche mB (1978Act33 B a ) 242, 248. [2 S ]CommitteAC n Environmentao e l Improvement Anal.Chem. 52 ( 1980) 2242. IT] KAISER, H., Anal.Chem. 4_2 (1970) 26A. ] WINEFORDNER[4 , ,D. LONG, G., Anal.Chem. _55_ (1983) 712A. [5] MOCAK,J., BEINROHR, E., Advances and Applications of Analytical Chemistry in Prac t ice--Trace Analysis, 6th Inter- national Conference, Rackova Dolina, CSSR, 3-6 June, 1986.

56 DISCUSSION

. KrivanV n youI .r talu relateyo k d blank o determinatiot s n methods woulI . d like to point out that the blanks, which without doubt determine the limits of detection and are in many instances also the dominating factor of the accuracy mainle ar , y introduce e predeterminatioth n i d ne stepth f o s analytical procedure. Therefore, these should be first discussed in connection with blanks.

. BraunT . Whee selectio comet th i n o data t s f ao n basr verificatiofo e f o n your statistical methodology don'I , t think journal papers coulgooa e db d choice. It seems that there are quite comprehensive data bases available for such purposes. f theo me e dat On coulth a e b collected e IAEth Ay b d Seibersdorf Laboratory during many years of intercomparison programmes.

R. Lindstrom doubI . t tha a bast datof e a exists (certainl publishednot y to ) test the usefulness of an approach such as you describe. One needs a large number of blank determinations by a stable procedure, and a body of calibration datr thafo a t procedure.

H. Ross. Conventional technique r leasfo s t squares fittin f dato g a would tend to "wash out" uncertainties in the slope in the low concentration region. Do you see this as a problem in the determination of Sc?

E.H. Klehr. t thia Yes st t sur,w bu pointsignifican no ho e m I' , t this valley is. We need to look at a carefully selected data base.

D.H. Smith. Can various definitions of limits of detection and quantification be evaluated by accessing and treating a large data base? I believe there mus mane b t y data base t industriaa s l labr EPA(o s , etc.).

E.H. Klehr. Yes d suc ,an thinI h , worso k k f greawoulo e b td benefit urgI . e that we consider not only the various limits of detection (quantitation, etc.) t variou,bu s way f analyzino s g precisio wells a n .

57 PLUTONIUM ISOTOPIC MEASUREMENTY SB GAMMA-RAY SPECTROMETRY VERSUS MASS SPECTROMETRY

R.L. MAYER, D.A. RAKEL Monsanto Research Corporation, Miamisburg, Ohio, United States of America

Abstract For several years Moun s activelha d y contribute e developmenth e o t th d f o t measurement of plutonium isotopic composition by gamma-ray spectrometry. This development effort has been conducted primarily in support of calorimetry and safeguards measurements s parA f .thio t s effort, Moun s participateha d a n i d numbe f samplo r e exchange programse Metalth r s Fo Exchang. e Program, Mouns ha d reporte e resultth d f o isotopis c measurements obtaine y b nondestructivd e ''gamma-ray spectrometry d destructivan ) e (mass spectrometr d alphan y a counting) means. Comparison e recenth f to s performanc f nondestructivo e d destructivan e e isotopics measurement t Mouna s d wil e presentedb l n additionI . e instruth , - mentation required for data acquisition and analysis of gamma-ray spectrometry measurements and their benefits will be presented.

59 DISCUSSION

D. Smith. In mass spectrometry, the 238 and 241 of Pu have interférants 238 and 241 Am; thus, mass spec analyses tend to be biased high. Mass spec analysis require r lesfo s s sample than gamma spectrometre th o s y, g) (u. s gv techniques are in many cases complementary.

R. Mayer e authorTh . s agree thae techniqueth t oftee ar s n complementarye Th . advantages of gamma-ray analysis is primarily the ability to obtain an analysis in a short amount of time while not requiring that the sample be opene r prepare o dy manner an n i d. particulaa Thi s i s r benefi n area f o n ai t interest, Safeguards. Mass spectrometr d alphan y a counting also have desirable features which gamma-ray analysis cannot provide, such as analysis combined with calorimetry to obtain a facility independent assay of material on site for a limited number of samples. Where problems arise, the item should be opened and sampled. The samples may then be analyzed by destructive means a late t a ,r date o verift , e calorimetrith y c assay (gamma-ra+ y calorimetry) value.

N .e gamma-ra Ikedath s I . y spectrometric method applicabl e low-leveth o t e l u isotopP e measurement n environmentai s l samples?

R.L. Mayer. First, we are not promoting the use of calorimetric (calorimetry and gamma-ray spectroscopy r environmentafo ) l samples e poinTh f .thio t s presentation was the comparison of gamma-ray and mass spectrometry results obtained at Mound through the Metals Exchange Programme.

In principal, the gamma technique, particularly the analysis program, could be applie o environmentat d l samples f course,O . e specificth f detectoo s r specifications, count times d samplan , e size woul de adjusteb hav o t e d with e measurementh t goal mindn i s . Remember though that measuremena thi s i s f o t relative isotopic composition e atoTh .m ratio 0 Pu/238 Pu/2324 23 s, , Pu 9Pu 9 241 Pu/239 Pu, and 241 Am/239 Pu are the quantities measured. To obtain a measure concentratioe sampleth th f n o ei ,u P som f o en alternative correction or comparison to standards is required. There are gamma-ray techniques for environmental measurements. It was not our intention to review these techniques.

60 URANIUM DETERMINATIO FISSIOY NB N TRACK COUNTING

F.F. DYER Oak Ridge National Laboratory, Oak Ridge, Tennessee, United State f Americso a

Abstract

To the analytical chemist, one of the most important aspects of nuclear tracks is the capability they give to quantitatively measure trace concentrations of alpha-emitting and fissile nuclide n matteri se procedurTh . r determininfo e g 235s i U simple in principle. A sample is packaged in a clean dust-free atmospher n contaci e t wit a trach k recorde d irradiatean r a n i d neutron flux for a suitable time. The recorder is etched to make the tracks optically visible and the 235U concentration is estimated froe tracth m k density relativ e tracth ko t edensit y produced by a 235U standard similarly treated. The detection limit for natural uranium is comparable to that of isotope dilution mass spectrometr d neutroan y n activation analysid an s s beeha n reporte beloe b o 0 picogramt dw4 r grame methope sTh . d can be applied to nearly all solids, including powders, and very little sample preparation is necessary.

INTRODUCTION

Sinc e discoverth e n 195i y 8 thae fissioth t n fragmentf o s U produce observable track n inorganii s c materials, studf o y

235 the phenomen f nucleao a r track n solidi s s toucheha s d many divergent areas of science and contributed to growth in knowledge in both the area touched as well as of the science of track formation. Development n basii s c nuclear particle physics, structur f o crystalline d amorphouan e s solids, geologica d lunaan l r datin f o mineralsg s diffusioga , n membranes, element mappin f fissilo g d alphan e a emitting nuclides, neutron flux monitoring d analyticaan , l chemistry determination e welth e l som f ar o sknow e n examples where important achievements have occurred e fascinatinTh . g stories behind most of the developments in the field up to 1975 are related in the superb reference work "Nuclear Tracks in Solids" by Fleischer Price and Walker (1) . Since 1977 the Journal of Nuclear Tracks and Radiation Measurements has been published to provide a focus for this area of research. Many of the current areas of research were described in 1983 at the 12th International Conferenc n o Nucleae r Tracks publishea r i d special issue of the Nuclear Tracks Journal (2).

Ia recenn t revie f highlo w y sensitive e methodth r fo s determination of uranium it was pointed out that only a few methods are capable of measuring the element in amounts of about

61 l nanogram and/or concentration e rangth f nanogram o en i s r pe s grad beloan m w (3). Isotope dilution mass spectrometry (IDMS) will detect picogram quantitie f naturao s l s i uraniu d an m perhaps currentl e mosth yt sensitive metho. Measurement ) (4 d s based on mass spectrometry have the advantage of being able to determine the individual isotopes of uranium. Resonance ionization spectroscopy (RIS d resonancan ) e ionization mass spectrometry (RIMS) are currently not as sensitive as IDMS but show promise of comparable sensitivity with further development (3). Conventional fluorometric methods will detect abou1 0. t nanogram of uranium, but measurements are not accurate unless the quantity of uranium is somewhat above 1 nanogram. Fluorometric methods base n laseo d r excitation have extendee th d sensitivity considerably (5).

Nuclear methods useo determint d e uranium include alpha counting, neutron activation analysis (NAA) , and fission track measurements. By counting for a 24-hour period the alpha particles from 238 0 unanogram3 , f naturao s l e uraniub n ca m measured with an uncertainty of about 30 percent. Neutron activation analysis can be carried out on both 238U, where either the induced 239U or its daughter 239Np are counted, and 2j5U, where one or more fission products are counted. The combined effects of the large thermal neutron fission cross section of 235U, the low isotopic abundance of 235U, and the small yield f fissioo s n products f naturao A NA ,l e resulth n i t uranium based on the activation Uo fbeing about a factor of

10 more sensitive than NAA based238 on 235U. The instrumental determinatio f abouo A picogram6 t NA y nb uraniuf o s r graf pe mo m high-purity silicon using 239Np countin s reportewa g d (6). Becaus e half-lifth e U (23. f mucs o ei h5) m shorter than that

of 239Np 239 (2.3 5 d) , it is practical to count a larger fraction of the paren e tdaughte th tha d f thuo nan rs obtai a highen r measurement sensitivit y b countiny g 239u. Althouge th h potential sensitivit r determininfo y g naturas li uraniuA NA y b m higher with 238U, it is frequently not possible to make use of this method (without chemical separations) because of interferences from other induced radionuclide s. Interferences are much more likely with the low energy gamma rays of 239U and Np than wite higth hh energy gamma rays emitte y somb df o e 239 the fission products of 235U. For example, only a few radionuclides, wita s 275sucN V photon* it s h4Ke a h , have gamma

rays with sufficien z t energie o interfert s e wite measurementh h t of the 1596 KeV gamma ray of the fission product 14CLa. A common technique to measure weak sources of high energy gamma rays in the presence of strong sources of low energy gammas is a thi e nt us opiec f leaw energo eo absor lo t d e y th bgamma t bu s transmit the high energy gammas. Thus instrumental NAA of uranium via the 235U isotope is useful in situations where NAA with the 238U isotope is not suitable due to interferences. The fact that the detection efficiency decreases with increasing gamma energy can be, however, another contributing cause for the lower measurement sensitivit uraniuf o 235A a NA vi mU f versuo y s

Althoug s possibli t i h o determint e e uraniue nanograth t a m m per gram level in certain matrices by the methods mentioned above with littlo chemican r o e l separations n mosi , t analyses

62 chemican a A l NA e cas separatiof th o e e uraniu n i th , f o or nm activation product s needei , d befor e e analysith b e n ca s completed. For example, the concentration of uranium in human urine can be determined by isotope dilution mass spectrometry at s 0.00a w level3 lo t onlug/ bu s ya s) Lafte(3 e uraniuth r s i m isolate d place a filamenan d n o d r evaporatiofo t d ionizatioan n n e masth sn i spectrometer.

o nucleaTw r methods, delayed neutron countind fissioan ) n(7 g track analysis t requirno o d ,e chemical separation o analyzt s e for uranium. Delayed neutron counting (DNC) is an activation analysis method based on the fact that several fission products with half-lives ranging from below a second to nearly a minute emit neutrons following beta decay e neutron Th e counteb . n ca sd with efficiencies of 5 percent or better and lower limits of natural 5 uraniunanograme rang2 th e measuredf b o n e i n m ca s . With a minimum of care, determinations of 235U can be effected instrumentally with no naturally occurring interferences. Fission track a methoanalysi t f activationo o d s i s n analysis, e fac s baseth i s tDNC ai n st o thad bu ,t when fissio 235f o n U occurs near the surface of a sample some of the fission fragments escape and produce latent tracks in adjacent track recorders. The track density, the number of tracks per unit area, is proportional to the uranium concentration which can be measure n nearli d l solidsal y .

DETAILS OF THE FISSION TRACK METHOD

According to Fleischer et al (1), nuclear tracks will only n fori insulatorm r o semiconductors s with electrical resistivities above 2000 Ohm-cm. Thus tracks wilt forn no li m silico t wil bu n nsilicai l . Track sn incompletela for y b m y understood mechanism that involves intense ionizatio n whici n h the atomic structur a microscopi f o e c e recorderegioth f o ns i r damaged. The damaged region can be observed with an electron microscope, but track densities are normally determined after the region is removed by chemical etching. Several methods, discussed below, can be used to determine track densities.

To determine uranium by the fission track method, a sample is packaged with a track recorder and the package is irradiated to a suitable neutron fluence (neutron flu x xtime) A standar. d containing a known amount of 235U and a track recorder are either included with the package or prepared and irradiated separatel n equaa r knowo o t ly n fluence A commo. n methof o d packaging samples is to tightly wrap them in aluminum foil. The uranium concentratio e unknowth n s i determinei n d froe th m relative track densitie e concentratioth d an s f uraniue no th n i m standard. Placin e recordeth g n intimati r e contact wite th h sample makes it possible to observe localized variations of uranium content ove e surface sampleth rth f o e , whereas allo"ir.g some distance betwee e sampl th nd recordean e r cause e emitteth s d fission fragment o spreat s t beforou d e recordeeth thet d hi yan r produce a more uniform distribution of tracks (1) . The more unifor e tracth m k distributio o determint e easies i th n e t i rth e average uranium concentration in the sample. The range of

63 fissio t irradiationbu n , s aboui fragmentcm r 4 t ai s n i hâvs e been mad n evacuatei e d quartz containers wite samplth hd an e recorder separated enoug o givt h a unifore m track density with highly inhomogeneous samples (1) .

Because uranium is present as a parts-per-million contaminan mosn i t t atmospheric dustsl necessars i al t o i ,d o t y sample treatmen d packaginan t a dust-fre n i g e environment such aa cleas n room wit a hlamina r flow clean hood. Fleischet e r al, indicated analyses can be made at 5 ppb, but that contamination problems generally limi e methoth t o uraniut d m r concentrationhighero b pp . 0 5 n Riley excellena f o sn i , t application of the fission track method to uranium measurement in semiconducto f clearo packagese nus hood e s ableth swa , y b , and "class 100" clean rooms, to measure concentrations as low as 1.5 ppb (8). The track density obtained by Riley on blank track recorders indicated that uranium concentrations as low as 0.04 e detectedb n ppca b .

Although many materiale b uses a nuclea dn ca s r track e mosth tf o frequentl o recordertw , ) y(2 s use r analyticafo d l measurements employing nuclear reactor irradiation e micd ar san a polycarbonate plastic, sold under the trade names Lexan (General Electric d ,Makrofo an USA), l (Baye , FRG)AG r. Mica track recorders are etched in hydrofluoric acid and the polycarbonate plastic e etcheH ar solutionss KO n NaOi dr o H . Detailf o s etching of most recorders are described by Fleischer et el. This type of polycarbonate is rather resistant to damage by the intense gamma radiation in nuclear reactors and can be irradiated in some facilities to neutron fluences of about 1017. Care mus e b taket o e keet polycarbonatnth p e cool during irradiatio r traco n k fading will occur. Experienc t ORNa e L with a pneumatic tube in the Oak Ridge Research Reactor with a thermal neutron 5 xlOs flushowf ha o 13x n that Lexas i n marginally useful, but that it cannot be used at all in a pneumatie Th e HFIcX 10) wher 1th *5 tub(6 e .fluR n s th ei ei x principaw lo lw cosd advantagelo an t s e plastiit th e f o sar c uranium impurity. The reader is referred to Fleischer et. al (1) and Enge (9) for details on the use of plastic track recorders. Mica is useful at much higher temperatures but suffers from the fact that samplee highesth f o st purity contai. U n b aboupp 5 t r measurementFo f verw o levels lo y f uraniumo s e verth ,y high purity fused synthetic silica used by Riley are recommended. Heraeus-Amersi f Sayrevilleo l . (USAJ ) . N , manufacturee th s silica and markets it under the trade name Supersil. The Supersi e fuse I gradlth df o esilic s essentialli a y bubble free uses nuclear i fo dd an r track recorders. Flame polishing wita h hydrogen-Oxyge ne fuseth torc df o hsilic a surface helpo t s eleminate defects that may be misidentified as tracks by automated scanning systemse .b e silicth Trackn ca n i a s effectively etched at ambient temperature in two minutes in concentrated hydrochloric acid.

Methods use o determint d e track density include simple manual counting under a microscope, computer controlled image analyzers, spark counting, electrical conductivity d lighan , t scattering techniques. These and other methods as well as technique o enhanct s e appearancth e f tracko e e discussear s y b d

64 Fleische . Usuallal t e rn simpl i y e optical counting, onla y limited fractio e fieldth f no s unde e microscopth r e countedar e , although some workers have counte e totath d l numbe f tracko r n o s the recorder (10). Spark counting makes use of plastic track recorders whose thickness (-10 urn) is less than the range of the particle that produces the tracks. The tracks are etched entirely throug recordee th h d sparksan r , that discharge through the tracks between electrodes placed on each side of the recorde e countedar r . Althoug e sparth h k counting methos i d rapid and accurate, it is limited to track densities of several thousand tracks per cm2 and in addition it would not be applicable to fission track measurements where reactor irradiation facilitie f o plasti se us prevenc e tracth t k recorders. Electrical conductivity methods alsf oo make us e thin plastic track recorder d measuran s e conductivity through tracks that have been etched through the recorder. Optical scanners, as was used by Riley, are suitable for track density measurement recordern i s s suc s fusea h d silica.

REFERENCES

) Fleischerd Walker (1 , "Nucleaan M. , , Price . B. L. ,R . . rP ,R Track n Solidsi s " Universit f Californio y a Press, Berkeley and Los Angeles, CA (1975).

(2) Special Volume, "12th International Conference on Solid State Nuclear Track Detectors" Nuclear Tracks and Radiation Measurements, 8 1-4 (1984).

, "Evaluatioal t e ) . F Dyerf Isotop(3 o n . F , e Dilution Mass Spectrometry for Bioassay Measurement of Uranium, Plutonium, and Thorium in Urine, NUREG/CR-3590, ORNL/TM- 9006 1984.

(4) Kelly, W. R. and Fassett, J. D. Anal. Chem 51 (1983) 1044.

) AnalKaminski, (5 al .t e Chem . ,R 3 (1981 5 . ) 1093.

(6) Dyer, F. F. et al, J. Radioanal. Chem. 22. (1982) 53.

(7) Dyer, F. F. et al, "A Comprehensive Study of the Neutron Activation Analysi Uraniuf o s y Delayemb d Neutron Counting, ORNL-3352, (1962).

) Riley (8 Anal. E . .,J Chem (1981. 53 . ) 407.

(9) Enge, W. Nucl. Tracks. 4 (1980) 283.

(10) Chakarvarti, S. K. et al, Intern. J. of Appl. Rad. and Isot 1 (1980.1 ) 793.

65 DISCUSSION

H. Ross. What other useful applications exist for track etch methods other thae analysith n f uraniumo s ?

F. Dyer. There are many examples of the application of nuclear tracks in Fleischer, Price and Walker's book Nuclear Tracks in Solids. One interesting applicatio measuro t s i n e verw nuclealo y r cross sections.

R. Rosenberg u expece ablyo b o analyzt eo o D t b uraniu. pp 7 ealuminiun i m m using delayed neutron counting without having serious problem e higth hy b s gamma-activity from the aluminium.

F. Dyer. We believe that we may be able to measure 3-5 ppb uranium in aluminium by DN counting because we have a lead shield between the sample and the BF detector.

V. Krivan. The topic of this meeting is comparison of nuclear methods with other methods. In this connection I would like to emphasize that the instrumental and/or the radiochemical NAA is at present the only one analytical technique allowing an accurate determination of uranium and thorium at the 1 ppb-concentration level and below.

F.F. Dyer e methoTh .f Resonanc o d e lonization Mass Spectroscop y sooyma e b n b levelpp f b abluraniuo ssu o determint e d e metalspresenn i an mth b t A pp e.t time it appears that fission track counting is the only method that will measure sucmatriw l levello al h n i xs materials. Conventiona A wilNA l l allow measurement of U at these levels in some material, e.g., but not many.

G. Revel. It is possible to obtain aluminium samples with a very low -9 -1 concentration in uranium: (C Uranium < 0,05.10 g.g ). This impurity y welive s l remove y zonb d e melting.

R.W. Bild. B, Li, N are other elements which can be determined by particle track techniques. The technique results in a map of elemental distributions as wel s concentrationsa l .

66 APPLICATIONS OF CERENKOV COUNTING: A USEFUL TECHNIQUE FOR ENVIRONMENTAL RADIO-ANALYSES*

J.S. ELDRIDGE Martin Marietta Energy Systems, Inc., Oak Ridge, Tennessee, United States of America

Abstract

Cerenkov radiation, produced by charged particles that exceed the velocit f a o particulalighy n i t r medium bees ha n, user somfo de quantitative estimates of radionuclide concentrations. In general, such applications have bee nspeciaw limitefe a lo t dcase s where th e radionuclid n questioi higa e hs beta-particlha n e energy majoe Th . r advantage of Cerenkov counting (the use of Cerenkov radiation for radio-analytica e lth simplicit purposes)n i f o sampl ys i e, preparation, and a secondary advantage is in the ready availability of automated counting equipment for performing the counting procedures. Ordinary liquid scintillation spectrometers available in most laboratories that conduct analyse wear sfo k beta emitter3 s H- suc s a h usee r C-1b o dy 4 withouma t modificatio Cerenkor fo n v counting. Sample preparation simply consist measurinn si 15-2ga aliquo0L m f to a colorless solution of the radionuclide into a suitable vial, followe y countinb d g wit a liquih d scintillation spectrometer adjusted for the feeble scintillations produced by the Cerenkov process. For 1- V betasme e settinth 2 , s e i similagth n o i t tha r3 tH- user fo d typical liquid scintillation cocktail. Deer harvested during managed hunts on the Oak Ridge Reservation have a potential bone burde f Sr-9no 0 from ingestio contaminatef no d water or vegetation from formerly-used solid waste storage areas neak Oa r Ridge National Laboratory. The 1985 hunts produced 926 animals that were a plastic-scintillato "screenedf o e us y b " r beta detector placed in close contact with an excised foreleg bone sample. This screening procedure e resulteremovath dee7 n i f rdo l froe harveso th t m e du t e detectioth f Sr-9o n concentrationn 0i s greater tha pCi/g0 3 n . From e th archived samples preserve a froze n i nd condition, Sr-90 concentrations were determined on a suite of 130 samples by use of a simple dissolution procedure on ashed bone followed by Cerenkov counting of the resultant solutions. The purpose of this study was to assess the accuracy of the screening procedure used in the field surveys and to compare the Cerenkov methodology with the contemporary radio-chemical method for Sr-90 in environmental samples.

Results showed that the Cerenkov technique provides a viable alternative to the more labor-intensive radiocheraical procedure. In addition, the results indicate a possible 4% inaccuracy in the ability to detect Sr-90 at concentrations greater than 30 pCi/g by the simple a plasti f o c e scintillatous r detector during field screening procedures.

* Research sponsored by the Office of Energy Research, US Department of Energy, under contract DE-AC05-84OR21400 with Martin Marietta Energy Systems, Inc.

67 DISCUSSION y countintr W.S u .e drie yo th g Lyon d d Di bon. e samples prio o dissolutiont r , etc. This would have taken a lot of time and trouble.

J.S. Eldridge. We did not try this technique because literature citations indicate dependenca d n granularityo y . Thus e woul,w do spen t havd dha e effort in grinding the ash to a given mesh size. It would save time if the ash were counted directly.

68 COMPARISON OF ISOTOPE DILUTION ANALYSIS WITH OTHER ANALYTICAL TECHNIQUES

J. TÖLGYESSY Slovak Technical University, Bratislava, Czechoslovakia E.H. KLEHR Oklahoma University, Norman, Oklahoma, United States of America

Abstract

Isotope dilution analysis e mos/IDAth ts i /e indicatousefu th f o l r radioanalytical methods, especiall r elementafo y l analysis w conceptNe . f o s IDA, suc s substoichiometria h d sub-superequivalencan c e IDA, have improvee th d sensitivity and expanded the applicability of the method for trace analysis. IDA has been used for the analysis of biological and geological materials, water, soils, semi-conductors, etc. In this paper, IDA is compared to other analytical methods, in terms of concentration range, limit of detection, selectivity, time and cost.

Isotope dilution analysis /IDA/ is the most often used indicator nuclear analytical method.

e basi e Th conservatioth c s principli A ID f activit o nf o e y upon dilution a radioactiv f I . e materia s dilutei l d wits non-radioactivit h e counterpart, the specific activity of the resulting dilution mixture is related to the original specific activity and the amount of material by means a simpl f o e conservation equation. This equatio e rearrangeb n ca n n varioui d s ways, depending on the conditions of the experiments to solve for the unknown mass of analyte in the sample.

The basic method and its many variants, has been used for the determinatio f traco n e elements /especially metals a wid n ei / varietf o y samples /metal, steal, alloys, rocks, minerals, water, soil, plastics, plants, biological materials, etc. d als/an o organic compounds /amino acids, steroids, vitamins, penicillins, insecticides, etc./ [1-6].

69 The e e usefumethofollowinb th n n i ca ld g general situations:

a substance is to be determined in a mixture of similar materials, but a quantitative isolation is impossible, the analyt w econcentration lo occur a t a s o thas , t losses sucs a h sorption onto vessel surfaces, etc., durin e separatioth g n procedures e inevitablear , the analysis must be performed quickly, e.g. because of decay or shifts in equilibria, the analyte is part of a large system, and only a part of it is available, e.g. water in a large living animal.

VariationA ID f o s

Classica A usee comparisoID lth s e specifith f o n c activita f o y radioactive tracer before and after mixing with a non-radioactive compound e determineb o whict s i h d /the analyte/ n otheI . r words e radioactivth , e tracer is "diluted" with its non-radioactive counterpart in the sample. The dilution causes a change in the specific activity of the added tracer, which can be measured and'used to calculate the amount or concentration of the component of interest in the sample.

A number of variations of classical IDA have been developed including:

Direct IDA r Singlo , e IDA, wher e non-radioactivth e e sampl s dilutei e d with a radioactive tracer, Reverse IDA, where a radioactive substance is diluted with a stable one, Derivative IDA, where the analyte is originally non-radioactive, but is made radioactiv a stoichiometri y b e c reaction wit a radioactivh e reagent, Double isotope dilution, where two radioactive isotopes of the same element are used, IDA after activation, where the radioactivity of the analyte is induced by an appropriate activation technique, Pseudo IDA n whic i ,e dilutin th h e samgth e analytt elemenno s i es a t e tracers s adequatelth ha t bu , y similar chemical properties.

Classica s bee e ha determinatio nth A useID r l fo d t leasa n t thirteen elements, and for analysis of various sample matrices such as water, soil,

70 plant d animalsan s . Separation procedures include precipitation, distillation, solvent extraction, coprecipitation, etc. Methods user fo d yield /mass/ determinations include gravimetry, colorimetry, fluorometry, flame photometry, etc.

e maiTh n disadvantag e classicath f o e l method e necessitth s i a s f o y second typ f measuremento e o thas , t specific activitie e determinedb n ca s . This limits the sensitivity of the method.

However e basith n ci , equation, onle ratith y f specifio o c activities occurs. The means that if exactly the same amount /or an exact multiple/ of materia e isolateb n ca l d from sampl d standardan e e equationth , s simplify drastically. This e mos concepe basith th t f s o oftesi t n use- d A methoID f o d substoichiometric IDA.

principlThe sub-superequivalencof e consist IDA epreparatio the in s of n o serietw f aliquoto se labelleth f o s d sample e serieon ;s isotopicalli s y diluted, the other is not. A constant amount of a suitable reagent is added o eact h e aliquoproducth d s separatei an t d countedan d e relativTh . e radioactivities of the isolated fractions are plotted versus the extent of isotope dilutioe intersectioth d an n f thio n s curve wit a linh e corresponding to a special condition is used to calculate the result.

e methoTh s similai d o substoichiometrit r n thai tA onlID ce th y activitie e portionth f o s s isolate e measurear d t differbu d sthan i fro tt i m the condition of isolating equal amounts need not be fulfilled.

Sensitivity

The sensitivity of IDA is limited by the following factors:

I/ The smallest amount which can be determined or purified in direct IDA, 2/ The original specific activity in reverse IDA, / Specifi3 c activit e dilutinth f o y g radioactiv e radioactiveth tracef o r o r e reagen n derivativi t e IDA, 4/ Equilibrium constants of extraction, hydrolysis, precipitation and similar separation reactions used in substoichiometric IDA, 5/ Blank values /reagent contamination/, 6/ Stability of reagents at low concentrations, sorption on surfaces, etc., in substoichiometri d sub-superequivalencan c e IDA,

71 7/ Volume f solutiono s s user substoichiometrifo d c IDA, 8/ Neutron, photon, or charged-particle flux in IDA after activation, 9/ Interferences.

The limit for the specific activity of the tracer is the case of carrier-free radionuclides, which is seldom possible in practice. In Table I hypothetical sensitivity ranges for the specific activity theoretically achievable are shown [9].

Factors 3-7 are typical for substoichiometric IDA, with 3 being the most important. It must be realized that this factor includes the counter efficiency and background. With automated procedures, the time available for e mosth countin te b importan y ma g t factor d sucan ,h procedure e generallar s y less sensitive than manual methods.

TABL. I E

Sensitivity Radionuclides

, 86Rb> 99Mo, 115cdi

181Hf) 182Ta> 187Wt 191OBf 203Hgi

, 21°Bi

, 72Ga, 82Br,

109Pd, 186Re

> 125sb) 137Cs

10-14 45Ca, 77As, 89Sr, 90Y, 95Zr,

, 131If 140Bap

72 e finitTh e value f equilibriuo s m constants explai y verw wh nlo y concentrations often cause difficulty in isolating precisely equal amounts of a species from two different systems. Large solubility products or small complex formation constants are good examples.

Blank values, rather than instrumental factors may limit the sensitivity f substoichiometrio c IDA .e reduceb Then y carefuca yb d l purificatiof o n reagents.

Stary [8] evaluated the sensitivity of good substoichiometric IDA -9 -8 -7 , corresponding 0 procedure e applicatio-1 th M 0 o 1 t g 0 s 1 a s f o n concentratio f substoichiometrio n c reagents. This sensitivity exceeds thaf o t activation analysis in some cases.

n examplA f greateo e r sensitivit f substoichiometrio y s comparea A ID co t d activation analysis is the determination of palladium. Whereas in the given case the smallest amount of Pd determined was 10 ng, the substoichiometric method was shown to be capable of determining down to 3 ng. Moreover, the latter metho s expectei d e applicablb o t y dtyp an f matrixo eo t e , whereas self-shielding effects were suspected to cause low results in activation analysis.

The detection limit of substoichiometric IDA is compared to other methods ir betteno s gooe Tablsee, a b as s drI n [10]i I eca tha thaA t .ID nI t other methods n favourablI . r superiofa es casesha r A detectioID , n limits /values are quoted to the nearest order of magnitude/. In Table III are presented some general characteristics of a variety of instrumental methods. In Figure 1 the applicability of IDA and other analytical methods are shown [2J.

Accurac d precisioan y n

By accuracy we mean the closeness of agreement between a true value and the mean result obtaine y applyinb d e samth ge experimental procedure th o t e same sample a large number of times. Since "tru" values are never known with certitude, statements of accuracy are never absolute.

Precision is a statement of the closeness of agreement among results obtaine y applyinb d e samth g e experimental procedur e samth e o t sample e several times. The smaller the random part of experimental errors, the more precise is the procedure. Some function of the standard deviation is commonly

73 TABLE II.

Detection limits of various analytical methods

Method Detection Volume of Limit of cone. limit sample f analyto e which mol. /S/ /ml/ wght of 100 /mol L-l/

Non-radioactive analytes

cfl -6 6 Ü Classical methods 10 10 10 H -12 -8 Ë Special ultramicro procedures 10 0.001 10 _CcJ u

-8 -7 ,1 — Absorption spectrophotometry 10 1 10 03 U -9 — 7 l-l Emission spectroscopy 10 0.01 10 H v -10 -10 .c u Inversion polarography 10 10 10 i -9 n) Mass spectrometry 10 0.01 10 u -13 -9 Gas chromatography 10 0.001 10 ^ -13 -10 D-, Atomic absorption spectrometry 10 0.01 10

11 10 SubstoichiometriA ID c io- 1 io- -13 -12 Neutron activation analysis 10 1 10

OJ dl Radioactive substances O D Z

-19 10 8 lonization and scintillation 10 1 io' -22 -21 Micro radiography 10 1 10

74 TABLE III.

Comparison of some instrumental analytical methods

Method A n a 1 y t e Cone .range^- Selec- Time^ Cost3 tivity

Classical pola- Iden.and detn.of 10°-l(r4 Good Short Low- rography about 80 cations medium of heavy metals; many organic substances

Inversion pola- Detn.of about 40 10-4-10-1° Poor Short Medium rography heavy metals

Atomic abs.spec- Detn.of many heavy 10-2-10-6 Good Short Medium trophotometrys 1 a t e m 5 Molecular abs . Detn.of practically io°-io- Fair Short Medium spectrometry l cational s except /UV, Visible/ alkali metals; a numbe f anionso r ; some organic species 1 IR spectrometry Idn.of functional io°-io- Fair Short Medium groups and detn. of structure of org. substances 1 5 Emission spec- Idn.and detn.of io- -io- Good Short Medium troscopy almost all elements /multielement method/ 3 Electron-raicro- Iden.and detn.of io°-io- Good Short High sonde elements with atomic no.>5; local surface analysis 1 1 NMR Spectrosco- Idn. ; structural io -io- Poor Short High py anal. ;detn.of mainly organic species

Electron spin Iden.and structu- 101-10~1 Poor Short High resonance spec- ral anal. of species trometry whicd h od hav n a e no. of electrons

Mass spectromet- Iden.of org. species, lOi-lCr1 Good Medium High ry structural ana.;rel. masses; iden. and detn of all isotopes ______t TABLE III. /contnd./

Methoe t dy 1 a n A Cone .range1 Selec- Time2 Cost3 tivity

Neutron activa- Multi-elemental Fair Medium- High tion analysis iden.and detn. ——-' long 7 Chromatography Iden.and.detn.of lo-Mo- Good Short Medium many organid an c inorg. species

X-Ray fluorescen- Multielement detn. 10-1-10-5 Good Short High ce with X-ray of elements with tubes at.n5 >1 o

X-Ray fluores- Multielement detn. 10 -10 Good Short Low- cence with radio- of elements with medium nuclide excita- at.no. >11 tion

Classical IRA Detn f organi.o c Fair- Medium Low and inorganic Good species which can be labelled with appropriate radionuclide

IDA- subs to ich io- Detn f mos.o t ele- Good Medium Low roetri d suban c - ments when appro- super equivalen- priate labelling ce methods nuclidee ar s available

Percentag f analyteo e ; sample weigh gram1 t .

2Short: seconds to minutes; Medium: hours; Long - days

3Equipment cost, Low 10,000$ < : ; Medium $ 10,000-75,000: ; High $ 75,00 > : 0

76 used to state precision values. In Figure 2 the precision of IDA and other analytical methods expressed as relative standard deviation s is shown and Figur 3 givee a comparisos e analysith f o n s time require r thesfo d e methods. This information can be used for the comparison of analytical methods, and for help in choosing a solution to a specific problem.

RANGE IDF APPLICATION 8 - 7 - 6 - 5 - A - 3 - 2 - 1 - 30 10 1 10 10 10 10 10 10 10 10 i i i i i i i

ACTIVATION ANALYSIS> t * IDA * > MASS SPECTROMETRY * > * > SPECTROPHOTOMETRY * > POLAROORAPHY

AA * D> FLAME PHOTOMETRY * > EMISSION SPECTROMETRY * 0 t> LASER * ^ MICROANALYSIS MÖSSBAUER SPECTROMETRY > » SPECTROMETR* Y TITRIMETRY * > XRF * * > GPVIMETRA Y

l i i • i • i • i • 'i io2 io'A io6 io"8 io"10 io"12 idtt id16 io"18 io"20

> OPTIMUM APPLICABILITY * TYPICAL APPLICABILITY

FIG.l. Range of application of some analytical methods

Most authors claim that substoichiometric IDA is precise and accurate. This shoul e takeb d n "cum grano salis" because this appraisemen valis i tr fo d cases wher a suitable e reagen d propean t r conditions have been chosend an , skilled operators are employed. Even in such a typical and well elaborated procedure as the determination of mercury with Zn dithizonate, the average of six single determination s 1.070+0.03uwa s g when 1.007iis takenwa g H g; that is, precision was within ~3% and accuracy not better than 6.3% [5].

77 COULOMETRY ORAVIMETRY TITRIMETRY POLAROORAPHY

SPECTRO- PHOTOMETRY AAS ACTIVATION ANALYSIS IDA XRF FLAME PHOTOMETRY MASS SPECTROMETRY EMISSION SPECTROMETRY

2 10' IQ' IQ' 10 100

FIG.2. Precision of analytical methods expressed with the standard deviation s .

1 235 10 20 30 50 100 200300 500 1000 mm i i l i i i nul i i i 11 i i i l i——l—i—l i i i il

ACTIVATION ANALYSIS

GRAVIMETRY

17///////////A— MASS SPECTROMETRY POLAROORAPHY AUGER AND PHOTOELECTRON SPECTROMETRY TITRIMETRY EMISSION SPECTROMETRY

ROENTGEN FLUORESCENCE SPECTROMETRY I SPECTROPHOTOMETRY FLAME PHOTOMETRY ATOMI S SPECTROMETRAB C Y

——————- ISOTOPE DILUTION ANALYSIS

COSR DETERMINATIOFO T E ELEMENON F NO T COST FOR DETERMINATION OF SEVERAL ELEMENTS -——- R DETERMINATIOTIMFO E N (mini FIG.3. Analysis timr differenfo e t analytical method-).

78 A positive exampl a highl f o e y accurate substoichiometric determination is the analysis for iron using solvent extraction with substoichiometric amounts of oxine. Kudo and Suzuki [7] found in this way 548+9 ppm of Fe in spinach whereas the certified /NBS/ value was 550^20 ppm.

Although substoichiometri s considerei A ID c d most usefu t traca l e element levelsn als ca e advantageouslb ot i , e precisyth user fo ed determinatiof o n major component /such as total amount or iron in silicate rocks amounting to e suiteb r routint fo dno y ema analyset I . / thin i s0 s e F cas f eo % 13 - 3 but coulusefua e b d l supplemen casen i t s wher e resultth e s obtaine y otheb d r e checkede b abov methodth o t en I e example. ar s e substoichiometrith , c 59 reagen s EDTd uncomplexewa tan A s retainewa a / d Fe n iroo d / n cation-exchange column. The precision and accuracy are 1.5% or better.

Advantages. drawback d trendan s s

e principaTh d decisivan l e possibilitth e s i advantag A ID f usino yf o eg non-quantitative isolating procedures, so that in some instances IDA is the only analytical method of solving a problem. This possibility allows the analys o perfort t n isolatioa m n quickly o choost , a purificatioe n method from a very wide selection of methods and to tolerate a partial decomposition of the substance analysed durin e analysisth g .

As the analyte is radioactive, its path through the analytical scheme can be followed, to be sure of its identity and to check for adequate purity by its constant specific activity. Also, certain types of IDA are readily amenabl o automationt e .

In IDA, there is no danger that the sought element can also be formed by a different nuclear reaction from another component of the irradiated sample /which may be a cause of false results in activation analysis/.

e substoichiometriTh c principl s bee e ha determinatio nth user fo d f o n more than 50 trace elements in pure substances, geochemical, biochemical and other materials e analysith n s i wel a f ,radioactivs o sa l e substances, sucs a h 6 - 4 - fission products with a detection limit of 10 -10 %, in special cases -8 . % 0 1 dow o t n

79 e trendTh s expecte e futurth n i ed developmen methoe th f do t probably includes : efforts to find new highly sensitive, chemically stable and selective reagents, elaboration of new effective separation procedures /solvent extraction, chromatography etc./, finding possibilities of simultaneous determinations of several elements, automation of the process for routine analysis.

n sub-superequivalencI e systemth A e samsID eth e studie f o typ s e a ear d those used in substoichiometric IDA. The main advantages of the sub-superequivalence method include:

higher sensitivity in systems where stability constants limit the sensitivity of substoichiometric IDA, systems can be studied where changing complex composition makes isolatio f equao n l amount f materiao s l impossible, onle phason y e nee e measuredb d .

The development of this radioanalytical method will probably involve the f othero e us , presently non-described systems, application f automateo s d version f procedureo s d applicationan s o othet s r samples /samplef o s environmental importance etc./.

Except for rare cases, IDA is not a multielement method, and it may not be practical to store radioactive isotopes of a large number of elements, or to prepare and keep a large assortment of labelled compounds which would be required for a multielement program. Moreover the availability of suitable nuclide d labellean s d compounds, particularly organic a species e b y ma , problem.

As compared to activation analysis, the major drawback in IDA is the necessity of using reagents freed from traces of the substance to be determined and the very low concentration level at which the isolation must be performed. Therefore IDA is not advantageous for the analysis of traces in samples that require large amounts of concentrated reagents for their dissolution, because their purification may be difficult.

80 One trend in the development of IDA can be characterised by a quotation from Lyon: "Ther o lacn e f informatioe o kseemb o t s n being passed bacd an k forth withi e literatureth n t continue i t mos bu ,f e rehashing o tb o t s , refurbishing or reevaluating of previously reported methods or techniques. We left our literature search with a feeling of "déjà vu" [11].

On the other hand, progress can be expected in some directions. A new e simplth possibilit s i e A determinatioID n i y e specifith f o n c activity using Cherenkov photometry. If the compound of the radionuclide is coloured or can be converted into a coloured compound, it is possible to measure the activity and the concentration of the compound. Kulcsâr et cl. [12] used, for the 32 determinatio e specifith f o ne yello cth activit, P w comple f o y x with molybdic and vanadic acids. The specific activity is determined from P counting efficiency, activite phosphoruth d an y s contene samplth f eo t usina g liquid scintillation spectrometer.

For the use of IDA in organic chemistry and biochemistry the increasing assortmen f commerciallo t y available labelled compound d theian s r decreasing price may be of great importance.

REFERENCES

1 TÖLGYESSY , BRAUN,J. , KYRS T. , "Isotop ,M. e Dilution Analysis", Pergamon Press, Oxford, /1972/. 2 KLAS, J., TÖLGYESSY, J., LESNY, J., "Sub-superekvivalentovâ izotopovâ zrieïovacia analyza", Veda, Bratislava, /1985/. 3 RUZIÖKA, J., STARY, J., "Substoichiometry in Radiochemical Analysis", Pergamon Press, Oxford, /1968/. 4 TÖLGYESSY VARGA, , J. , "NucleaS. , r Analytical Chemistry", Veda, Bratislav Universit- a y Park Press, Baltimore, /1972/. 5 TÖLGYESSY , KYR§ ,J. , "Radioanalytica ,M. l Chemistry", Veda, Bratislav- a . HorwoodE , Chichester n press//i , . 6 TÖLGYESSY , KLEHR,J. , E.H., ZACHAROVA LESNY, . Z ,, KLAS J. , J. , "Isotope Dilution Analysi f Environmentao s l Samples", IAEA Consultants' Meetin n Advanceo g Nuclean i s r Technique r Analysifo s f Environmentao s l Samples, Karlsruhe, FRG, 12-21 February /1986/. 7 KUDO SUZUKI, K. , , J.Radioanal.Chem.,N. 9 /19805 , / 605.

81 8 STARY, J., "Modern! radioanalytické metody /Modern Radioanalytical Methods/" in "Nové smery v analytické chemii", ed. Zyka, J., SNTL, Prague /1983/. 9 GUINN, V.P., LUKENS, H.R. Jr. "Tracn ,i e Analysis, Physical Methods", G.H. Morrison, éd., Interscience Yorw ,Ne k /1965/. 10 MAJER, V., "Zâklady jaderné chemie", SNTL, Prague, /1981/. 1 1 LYON, W.S., Anal.Chero. 4 /19825 , / 227R. 12 KULCSAR, F., TEHERANI, D., ALTMANN, H., J.Radioanal.Chem., 63 /198l/ 217.

DISCUSSION

G. Revel havI o comment. tw e makeo t s :

The first concerns the comparison between the different methods of analysis. Table give y Profb n . TolgyessJ . e verar yy interesting because they situate the possibilities of these methods. However, it is important to consider that all methods do not oppose each other but are complementary. In practice, the choicn analyticaa f o e l method depend n severao s l factors e e naturth th :f o e sample and of the impurities, the concentration range required and also the f futuranalyticao e us e l results r exampleFo . materian i , l studie e choseth s n analytical method is not the same for a quality control and for a correlation betwee properta n e materia e concentratioth th f r o yd severao an le on l f o n selected impurities. y seconM d comment concern e limitth s f sensitivityo s a larg w eNo .numbe f o r methods have very low detection limit, but in fact for concentrations lower than one p.p.m., it is often the blank values which limit the detection in real samples and not the potential sensitivity of the analytical method.

82 THE CONTRIBUTION OF RADIOTRACERS TO THE DEVELOPMENT OF TRACE ELEMENT ANALYSIS

. KRIVAV N Sektion Analyti Höchstreinigungd kun , Universität Ulm, Ulm, Federal Republi f Germanco y

Abstract This paper discusses application e radiotracerth f o s o t s methodological studie tracn i s e element analysis. Owing some unique features radiotracee th , r techniqun a s prove e eha b o dt indispensable means for examining the individual steps of an analytical procedure and revealing the concomitant sources of systematic error. mosn I t cases radiotracee th , r technique th s ei optimum approach somd an , e problems canno solvee tb othey db r means examplee Th . s discusse morn di e detail were taken from the analytical chemistry of mercury, sulphur, lead and selenium. The radiotracer investigations aimed at improving the accuracy of analytical procedures.

1 INTRODUCTION 2 CHARACTERISTIC FEATURES OF THE RADIOTRACER TECHNIQUE [5,6] Reliable analytical dat n traco a e element contentf o e ar s principal importanc mann i e y field f scienceo s , medicined an , e radiotraceIth n r technique e componenth , t whose behavior technology The analytical techniques for trace element analysis during a given process is to be investigated, is labeled with a may be either direct instrumental or multistage techniques radioactive isotope This is usually achieved by adding a involving chemical processin e samplth f o ge e prioth o t r radiotracer containin gmeasurabla e amoun f radioisotopeo t s sa measurement stee multistagTh p e techniques generaln i ,e b n ca , atoms, ions or molecules to the test sample The basic assumptions take methodologe modea th s na r fo l f tracyo e element analysis in using radiotracere sar Typically, an analytical process consists then of the following steps sampling, storag d pretreatmenan e e sampleth f o t , decomposition, separation (preconcentration), and measurement e rat f decaeTh o 1 y measure proportionas di numbee th o t l f ro atoms left undecayed Each of these steps can be a source of systematic error In general mose th , t significant source errof s o contaminatione rar s 2 The relative proportions of the radioisotopes and stable and losses The blank not only determines the limit of detection, isotopes of an element are not changed by the processes under but with decreasing trace element concentratio t becomeni e th s investigation predominating factor of accuracy Considerable losses of the 3 Radioactive radiation doe t affecno s t other propertief o s trace elemen f intereso t n occu ca ty adsorptiob r n and/or labeled matter volatilization during the analytical process However, many other sources of error are also possible Assumptio derives i n1 d fro theore mth f radioactivyo e deçà) and, within the counting errors, can be considered as fully true The radiotracer technique has proved to be an excellent means for examining the individual steps of the analytical process and revealin concomitane gth t source f systematio s c error [1-4n ]I In practice, three factor determinn sca welw eho l assumption2 many instances, it provides the best approach, and some problems is fulfilled the punty of the radiotracer, the quality of labeling canno e solveb t y otheb d r means e Wit th e hel th f ho p and the extent of isotope effect Unsuspected radionuchdic and'or radiotracers, many established analytical techniques were radiochemical impurities can behave otherwise than the optimize larga d edan numbepowerfuw ne f o r l techniques wa s radiotracer durin e procesth g s studie d alte e isotopian sth r c develope e radiotracerTh d s have greatly contributede alsth o ot proportion r perfecFo s t labeling e radiotraceth ,e th d an r solution of many general problems of trace element analysis, such componen e testh t f samplo t e traceb same o et th d en musi e b t s determinatioa f differeno n t analytically relevant constants, chemical form In order to fulfill this requirement, in man% cleanfication of reaction mechanisms, electrode processes etc In instances special labeling techniques must be used such as in situ addition, several important determination method basee sar n do labeling by activation of the material to be examined or in VISQ utilizatioe th f radioisotopeno s metabohzation Incorrect labeling can be a reason for serious systematic errors of the radiotracer technique The isotope effect In this paper, the potential of the radiotracer technique for results fro face mth t tha behavioe tth f differenro t isotopet no s i s studyin individuae gth l step f tracso e analytical procedured san exactl same yth e becaus f theieo r differenc masn i e s This effect identifyin sourcee th g f systematiso c erro discusses i r d dan mus consideree b t case th f hydrogen eo n di r othefo , r elements demonstrate selectey db d examples the error due to isotope effect is at or below 1 %

83 Assumptio t exactlno t e casis y th f 3 n o fulfillede n I o to , The radiotracer technique enabled to clarify the biological substances with high specific activities, chemical changee th f so methylatio f mercurno pury yb e culture f microorganismso d san radiotrace radiatioy b r n effect n givca se ris o considerablt e e by mixed cultures of environmental systems such as lake water systematic error, and must therefore be considered and sediments [7-9], fish intestinal tract [10] and other systems In several cases, however, methylmercury detecte e coulb t dno n di The main advantage of the radiotracer technique is the the system studied where it should have been present possibility of direct detecting the characteristic radiation in any Experiments using 203Hg provide n explanatioa d e th r fo n stage of the analytical process with rapidity and simplicity absence of methyJmercury it was demethylated to volatile Another important featur radiotracee th f eo r techniqu higs it hs ei elemental mercur ] Furthery [7 s founwa t di , that some bacteria sensitivity Depending on the half-life of the radioisotope and the such as Eschenchia coll, Skaphyloevecus aureus, several species ability to determine its actual specific activity in earner-free f Pseudomonaso d soman , e alga n reducca e e ionic inorganic state, picogramm amounts and less of the component of interest mercur elementao yt l mercur r examplFo y e [11], when s soiwa l can be traced through the analytical process The weights of some amende l|io dt g Hg/g soil with mercury(II) nitrate labeled with earner-free radionuclides having an activity of 10 Bq, which, in 2 203 general, can still be counted easily, are given as examples in °3Hg, 20-45e th Hg-activit f %o y were lost from different Table I However, for different reasons, the real earner-free soil types dunng the first 6 days The volatility rate decreased to a concentration alwaye ar s s higher tha e theoreticanth l Further, minimum after about 100 h Steam-stenlized soil lost only about the radiotracer technique has a high degree of universalit labelee th f yo d mercur% 2 ensuro T y e thamercure th t y loss swa Radiotracers are available for almost all elements Only the initiated biologically, the sterile soil was inoculated with the havl A ed exclusivelan e N , O , yN , verelementB , y Li short , sHe - respective nonstenle f soiincubationo lh afte0 16 rwithid an , n8 h, the inoculated soil started volatilizing mercury, reaching a lived radioisotopes that are not suitable as radiotracers maximum after 22 h Table I These examples show that the contents of the individual Weight Amounts of Carrier-Free Radionuclides mercury compounds as well as of total mercury can be seriously Corresponding to an Activity of 10 Bq affected by bacterial conversion processes Therefore, these processes mus consideree b t botn di analyticae hth l procedurd ean the interpretation of results Amount y weighb s t a t Radionucllde '1/2 10 Bq

NaCl HC1

3 SELECTED APPLICATIONS The investigation of the efficiency of this sampling device as a The broad applications of radiotracers to investigations of function of temperature using sulphuric acid labeled with 35§ analytical procedures include especially the following topics showed that more than 99% of the sulphuric acid was collected in the dénuder at 413 K The effect of a number of salts on the - Presamplmg factors collection of H2SO4 was examined in similar manner - Sampling (efficiency, contamination) Sampl- e homogeneity 3.3 Homogeneity of spruce needle samples - Sample storage (losses, contamination, chemical changes) In sampling bulk material fundamentae th , l requiremen thas i t t -Sample drying the proportion of the component of interest is, within the limit of (losses, contamination) error, the same in sample as in the whole Usually, subsampling is - Sample decomposition earned out on the sample to obtain proportions required for a (yields, chemical form, losses, contamination) single analysis replicatr fo , e analyses analyser fo r ,o differeny sb t - Separation and preconcentration techniques In subsampling, the degree of homogeneity is of (separation factors, yields, distribution coefficients, kinetic pnncipal importance if precise and accurate results are to be studies, studies of mechanism of the separation processes) obtained However, the homogeneity requirements can also - Determination stage depend on the determination method, as for different methods (behavior of the element determined, influence of various different sample portions are used for analysis parameters) For example, for the analysis of spruce needles by INAA and AAS, respectively, samplmg 0 e 25 portion d ,an g f abouso m 0 7 t Some selected example discussee sar morn di e detail below requiree ar d Large vanation elemene th f so t content needlen si s Pre-samplin1 3. g behavio f mercurro y same oth e werf eag e foun r differenfo d t branche same th f eso whorl, and even for different positions within one branch [13] The toxicity of mercury strongly depends on us chemical state Therefore e preparatioth , a representativ f no e samplf o s i e which shoul e knowdb n befor analysise th especiee th s f a , so principal importanc e radiotraceTh e r techniqu s user wa edfo concer determiny ma n e samplin , gsampl e pretreatmend an t examinin e e courshomogemzatioth th g f o e n procesr fo s storage as well as other steps of the analytical process Basically, subsampling examplar e elementh r fo y t manganes r thiFo se mercury occur environmene th n si elementas a t l mercury vapor, purpose, some whole needles were labeled in situ with ^°Mn by and as inorganic and organic mercurials These different species irradiation in nuclear reactor The irradiated needles were then mus differentiatee b t d with regar theio dt r input, transpord an t mixed with umrradiated approximately 1 50 The needles were mutual conversion then homogenized by the brittle fracture technique with the

84 Mikro Dismembrator (Braun e homogeneizatioTh ) n standard Investigatio4 3. f preatomizationo n behavio graphie f leath r o n di c deviatios wa d ~25 r seveg an fo nm 0 g nm aliquot0 ~7 f o s furnace determined by counting the ^6^jn activity The results proved ) tha(Fi1 gt acceptable degre f homogeneiteo obtainee b n yca d Valuable information can be obtained by the radiotracer for both sample portions during a homogemzation time of at least technique also about the behavior of the elements in the graphite 5 mm furnace dunn e individuagth l temperature stepth f so e program Such studies often reveal sources of systematic error and facilitate the optimization of the determination procedure The potential of the radiotracer technique m this field is demonstrated by usin gexampln leaa s da e [14experimente ]Th s were executed as follows an appropriate portion of a sample labeled with 203p5 was introduced into the graphite tube, and the tube was counted The e tubs placenth wa e d inte selectee furnacth oth d an ed temperature program started After the step to be studied had been completed e temperaturth , e progra e s stoppeth mwa d dan tube counted agai e graphitTh n e furnac places closea wa e n di d Plexiglas chamber, whic s connectehwa exhausn a e o th t d f o t radiochemical laboratory classB The results for the thermal pretreatment are illustrated in Fig e seeb n ca thae maximut I th t 2 m tolerable pretreatment temperatur mainls ei y dependen concentratioe th n o t chloridf no e samplee inth , chloride promote lose f dth leaso d (Fi ga,b,c2 ) Any significant improvement can be achieved by using the L'vov platform (2 d,e) The pretreatment temperature can be increased additioy b hydroge% 1 f no argoo nt g,hadditio2 y n( b r )o f no ammonia to the sample solution (2 i), the stabilization is probably 5 1 0 1 S due to formation of volatile hydrogen chloride and ammonium t(mm) chloride, respectively, that remove the chloride from the matrix Fig I beforn interferca t i e e with lead Using ammonium dihydrogen Dependenc e homogeneitth f o e y standard deviation e {n«7th n o ) phosphat as matrix modifier, the pretreatment temperature can t-omogenlzatio differeno ntw tim r fo e t sample portions (-7g n 0 (Fibese C ° th a,b gt3 0 bd stabilizatioe)40 an increase- 0 20 y db n and -250 mgl of spruce needles homogenised by the brittle frac- of lead is achieved with ammonium dihydrogen phosphate as the ture technique using the Mikro-Disaiembiator II modifier combined with usin L'voe gth v platform (Fi c,dg3 )

HNO3 O01 M CI 1%

• Unn1 « • Urine 1 A Ui SI

o so

• HNO] 001 M • MCI IM l NaCI 1% NaCI 1% • Wilh NHi

• Without NH4

RAMPTIMC (Si Fig 2 Influenc e ramth e pf heatin o th etim n * e g step from charring to atomization terrperature for various matrices' (a-d) graphite tube (e,f v platforvo )L m (g,hl graphite tubd additioan e f o n NH^HjPO ) graphit(i ^ e tv.be, solid samples (k,l) graphite tube and 11 hydrogen added to the flowing gas (m) graphite tube and addition of ammonia

85 z A.

z • MNO3 001 M 5 • HCI 1 M H U NaC« % 1 I

W SO S • Blood it Urin• » t

-l——L. O 9O 110TOO O 0SO 13O110 O TO 0 13OO

PRETREATMENT TEMPERATURE |«C) Fig. 3 Dependence of the losses of lead en the pretreatrrent temperature for various matrices to which NH.H„PO, v.as added (a,b) rraphite tube, (c,d) L'vov platform

The heating phase from the pretreatment temperature to the 3.5 Error-diagnostic investigation of selenium determination bv atomization temperature can also give rise to systematic errors, hydride generation AAS especially when the absorbance is measured in peak heights During the heating phase, lead can evaporate as lead chloride Losses of selenium up to 50% were reported in the which is not detected in the absorption measurement This heating determination of Se m wastewater and other environmental phase cannot be studied at all by absorption measurements samples by hydride generation AAS involving treatment with H2O2andHCl[15] For the gas stop mode, the losses of lead with and without addition of modifiers were investigated at different heating rates With Se-75 as the radiotracer, all the essential steps from a chamng temperature of 500°C (without phosphate) or (mineralization by sulphuric acid and hydrogen peroxide, 700°C (with phosphate) to the atomization temperature of reduction of Se(VI) to Se(IV) by 5 M hydrochloric acid, hydride- 2100°C usin e gas-stogth p mode result Th eillustratee ar s dm generation, hydride transport) were systematically examined Fig 4 as the dependence of the Pb-203 activity remaining in the [16] Non f theseo e step responsibls swa seleniue th r efo m losses graphite tube on the ramp time Again, the losses of lead are Preliminary radiotracer measurements indicated e thath t strongly influenced b> the chloride concentration of the matrix recoverie f seleniuso hydride th mn i e generation procedure could (Fig 4 a,b) The lo\ver losses with l M HCI as compared with 1 % be influenced by the time elapsed between the reduction of Se(VI) NaCI solutio explainee b n nca alreadI escapy db HC f eyo during to Se(IV) by hydrochloric acid and the generation of hydrogen the drying and pretreatment step selenide Additional more detailed experiments showed, that the Se 75 activity in the hydride-generation flask increased with A substantial reduction of lead losses could be achieved with the increasing times between the Se(VI) to Se(lV) conversion and the L'vov platform, these results are in good agreement with the hydride generatio seleniue th f no mEight% remaine95 o yt n di theory of delayed heating on the L'vov platform Ammonium the hydride generator, when hydrogen selenide was generated dihydrogen phosphate effectively stabilizes lead also during one week after the Se(VI) to Se(IV) conversion We were able to heatin atomizatioe th o gt n temperatur lealoss o n edtwa froy man prove that this increase of not reducible selenium with time is of the matrices after one second ramp time, but strongly matrix- cause sloa y wdb back oxidatio f Se(IVno Se(VIo t ) chloriny b ) e dependent losses occurred at longer ramp times Significant formed from hydrochloric acid and hydrogen peroxide The stabilization can be achieved by addition of hydrogen to the redox reaction between selemte and chlorine sheath gas (Fig 4 k,l) or by addition of ammonia to the solution ) Alsm (Fi phosphate thin 4 oi gth sf caseo e eus , modified an r L'voe th v platform lead higheso st t degre stabilizatiof eo f leano d and minimizing the matnc effects (Fig 5) is an equilibrium reaction Using standard heats of formation in 1 M aqueous solution an enthalpy value for this reaction, To investigate errors associated with the analysis oe equilibriumf Th soli obtaines i C ° d1 5 l 2 t AH d,mo a f o -14J , k 49 samples, radiotracer experiments were S earne NB witt e dou h th which favors selenic acid at room temperature, is shifted toward standard orchard leaves (SRM 1571) and bovine liver (SRM selenous aci t highea d r temperatures Radiotracer experiments 1577) The different behavior of lead during the pretreatment showed that, in equilibrium, only 5% of the selenium are present and heating to the atomization temperature can be explained by as selenate at 100°C whereas 97% are in this form at room the different phosphorus/lead ratios (10500 Hg/g P, 0 35 ^ig/g Pb temperatures in bovine liver, 1980 orchar n i ng/u.g/6 4 b , gP g P d leaves) High phosphorus/lead ratios stabilize leasimilaa dm r manne s doea r s Fig 6 shows the decrease of the selenium(IV) concentration addition of the matrix modifier ammonium dihydrogen with time reducnoth ef o pas d nt en afte wite th hr hydrochloric phosphate Radiotracers could also provide information aboue th t acid Twenty minutes after the reduction 93% of the selenium movemen f o leat d withi e graphitth n e tube durine th g e tetravalen s th stil n wa i l t form Afte weee selemum(IVe on r kth ) temperature program concentratio decreased solutioA nha % 6 selemtf no o M dt 5 n ei

86 100O 9O 0O BO O TO 0 60 O SO 1ÖOO O BO O BO O TO O 6O O SC 0 10O90 O O BO 0 70 SC0 60 O

PRETREATMENT TEMPERATUR) C I* E 4 LosseFig. f leaso d froe graphitth m e tub. pretreatmenvs e t tempera- turr varioufo e s matrices: (a-c) graphite tube; (d,e) L'vov platform ) graphit(f ; e tube, solid samples; (g,h) graphite tube hydroge% 1 e flowin d th an n l graphiti n (i g gas; e tub d addian e - tio f ammoniao n .

. • M N Oj 0 01 M M I • MCI :: ANlCI 1%

4 3 123 ! 1 4

RAMPTIME (i| Fig. 5 Infljence of the ramp time in the heating step from charring to atomization temperature for various matrices by using the L'vov platform and NHHPO as the matrix modifier

100 -

S «o

0 25 SO 7S WO Ui 1JO* TIM) (h E 6 FigDecreas . e SellVth f Io e concentration with time afte e reductioth r f o n Se(VI o Se(IVt I ) with hydrochloric acid

87 hydrochloric acid, whic d beehha n saturated with chlorin] [5 t a e KRIVAN, V , ' Characteristic Features of Radioactive 25°C, contained only 3% Se(lV) ater 2 days Hydrogen peroxide Substances and Principles of Experimental Techniques , in the absence of hydrochloric acid did not oxidize selenite to Nuclear Analytical Chemistry I (TÖLGYESSY.J , selenate VARGA S , KRIVAN, V , Eds), University Park Press, Baltimore (19715 13 ) From thes concludee b result n ca t i s d that hydrogen peroxide oxidizes hydrochloric aci chlorino dt chlorined t 25°a ean , n i , [6] BOWEN , "Introductio M J H , Radiotraceo nt r turn, oxidizes selenit selenato t e e Selenat t reduceno s i ey db Experiment , Treatiss Analytican eo l Chemistry, Par, I t sodium borohydnd e amounth d f seleniuo an te m hydride Vol III, 2nd Ed , (KOLTHOFF, I M , ELVING, P J , reaching the atomization tube decreases with tune The oxidation KRIVAN , Eds) V , , Wiley Yorw Ne ,k (19863 19 ) of Se(IV Se(VIo e preventet )b n ca ) flushiny db chlorine gth e formed during the reduction of selenate in boiling hydrochloric [7] SPANGLER , ROSE J MILLER, W M , , J , H H , acid from the solution with a stream of nitrogen Recoveries of Science 1£Q (1973) 192 98% independent of time up to three weeks between the reduction of selenate and the generation of hydrogen selemde were obtained [8] BECKERT, W F , et al, Nature 242 (1974) 674 for selenium in standard solutions and in wastewater samples, if chlorine was removed by flushing with nitrogen ] [9 FURUTANI , App , RUDD M A , l W Enviro J , n Microbio (19800 4 l 0 )77

REFERENCES [10] RUDD, J W M , FURUTANI, A , TURNER, M A , Appl Environ Microbiol _4Q (1980) 777 [1] KRIVAN "Methodologica, V , theoreticaland l Studiesin Analytical Chemistry Using Radioactive Tracers", [11ROGERS] , R D , McFARLANE, J C , J Environ Anal Nuclear Analytical Chemistr (TÖLGYESSYI yI , J , 2(1979)255 VARGA , KRIVAN S , Eds), V , , University Park Press, Baltimore (1972 5 36 ) [12] NIESSNER, R , KLOCKOW, D , Intern J Environ Anal Chem 2(1980) 163 [2] McMILLAN, J W , The Use of Tracers in Inorganic Analysis", Radiochemical Methods of Analysis, Plenum, [13] KRIVAN, V , SCHALDACH, G , Z Anal Chem 224 New York (19757 )29 (1986) 158 [3] KRIVAN, V , Talanta 22 (1982) 1041 [14] SCHMID , KRIVAN W , , Ana V , l Che J (1985m5 0 )3 [4] KRIVAN, V , "Application of Radiotracers to [15] WELZ, B , MELCHER, M , Vom Wasser_£2(1984) Methodological Studies in Trace Element Analysis", 137 Treatise on Analytical Chemistry, Part I, Vol III, 2nd Ed (KOLTHOFF , ELVING M I , , KRIVAN J P , V , [16] KRIVAN , PETRICK V , , WELZ K , , MELCHER B , , Eds), YorWilew kNe y , (1986 9 )33 , Ana M l Chem 5J(1985) 1703

DISCUSSION

W.D. Shults. Your wor Cr(IIIn o k Cr(VIs )v extremels )i y importantu yo o D . plan to publish it?

V. Krivan. Yes, it is in press in Analytical Chemistry.

88 APPLICATION OF THE ISOTOPE EXCHANGE METHOD ANALYSIF O SPECIATIVE TH O ST E DETERMINATION PHOSPHORUF O ENVIRONMENTAN SI L SAMPLES

N. IKEDA Department of Chemistry, Tsukuba University, Sakura-mura, Ibaraki-ken, Japan

Abstract

3 3 Speciative determination of phosphorus (P04 ~, P03 ~ and naturan totai ) P l l water samplee achieveb n ca sy usin b de th g

reverse isotope exchange method. Five ml of a natural water sample is taken and dil. HaSOi,

solutio f ammoniuo n m molybdophosphat definita d an e e activity

32 3 (ae) of n POit(i ? form e added e solutio)ar Th . s shakei n n with the 1,2-dichloroethane solution of tetraphenylarsonium molybdophosphate which contains a definite amount of P (K'i).

After the quickly attainable exchange equilibrium between phosphate in the aqueous phase and tetraphenylarsonium molybdophosphate in the organic phase, the activity of the organic phase (a measureds 2)i e amount sample Th P (Wth . f n xo se)i can be obtained by the equation of Wx = ( a^/az - 1 )Wj.. In this condition, only P0i»3~ exchanges, while P0a3~ dose not. Therefore ,e determineb POi« n e presencca 3"th a PO n i df o e and unexchangable phosphorus.

When the sample water is treated with Br water in advance, 2

3 3 the amount of POi, ' + P03 ~ can be obtained. When the sample

is treated with HSO^ and HN0, then the amount of total P can 2 3 be obtained.

e methoTh s comparei d d wite conventionath h l spectrophotometric method (the molybdenum-blue method). This methos i d

89 much simpler and rapider, and moreover, the speciative determinatio possibles i n .

1. Introduction

Among various method f radioanalyticao s l chemistrye th , isotope exchange analytical method (IEA method) seems still rather unfamiliar, and the practical application of the method

is still remained to be developed.

Langer [l] determined the amount of silver chloride precipi- tat y exchanginb e g with radioactive silver ion. Since then, the isotope exchange determinatioe methodth r fo s f silveo n , [2 r mercur, 3] y [4], iodine [5], bismuth [6], antimony [7], arsenic [8], thallium [9], phosphorus [10,11], etc. were proposed. The method used in these literatures is, according to my

classification, the direct isotope exchange one, in which the

inactive specie s determinei s y exchanginb d g wite radioactivth h e

reagent n thiI s. case, therefore s necessari t i , preparo t y e the radioactive reagent in advance. The author [5] proposed

the reverse isotope exchange method n analys ca y whic, b e on eh

the radioactive sampl y exchanginb e g wite inactivth h e reagent. The reverse method is also applicable to the inactive sample when the sample is spiked and made radioactive with the carrier-

free radioisotope of the species in question. In this work, the application of the reverse isotope exchange method to the determination of phosphorus in the environmental water samples is shown comparing the results with

those obtaine e conventionath y b d l spectrophotometric method.

90 The method is further applied to the speciative determination of phosphite and total phosphorus in the environmental water.

2. General principl e isotopth f o ee exchange methof o d analysis 2.1 Direct isotope exchange method

To the sample solution which contains the inactive species

e determinedb o At X definita , e e radioactivamounth f o t e species BX* (weight of X: W]., activity: ai) is added and mixed well. After the isotope exchange reaction

X B + * AX i ;= * BX f - X A proceeds completely, AX and BX are separated from each other,

and the activity (a2) of the BX fraction is measured. e exchangAth t e equilibrium e radioisotop,th X distri f o e -

butes homogeneously among AX and BX and the specific activity

becomes the same, so that the following equation holds good:

ai a.

X specieA e determinedn b i e amoun o X t sth f s whero 3 ti W x .W e is readily obtained as follows:

Wx = ( ~ - 1 ) Wi (1)

2 2. Reverse isotope exchange method The reverse isotope exchange method is useful, when the

sample to be determined is radioactive and its activity is

measurable.

A definite amount of inactive BX (weight of X: Wi) is added

e sampltth o e solutio f radioactivo n , x W * (weigh : AX eX f o t

91 activity:ao). After the isotope exchange reaction attains equilibrium, both specie e separatear s d from each e otherth d an ,

activit fractioX B ye (a th measureds i nf 2 )o . e exchangAth t e equlibrium e followin,th g equation holds:

ao a2

Wx + K! Wi Therefore,

Wx = ( — - 1 ) W (2). 3-2

The reverse method can be also applied to the inactive

sampl e direcn placi th e f o et method n thiI s. mads casei eX ,A

radioactive, in advance, by adding the radioactive AX species

e negligibl(activityth f o ) 0 a :e weight. This n methoa s ha d advantage to the direct one that tedious preparation of the

radioactive reagent BX* can be eliminated.

3. Determination of phosphate ions

Determination of phosphorus in the environmental water is

important froe standpointh m f wateo t r pollutio n additioi n o t n

e geochemicath d agriculturaan l l importance e simplTh d . an e

rapid method of analysis for phosphorus is, therefore, very much

desirable. Perezhogi d Sidorovan n a [12] pointe t thaou dt rapid isotope exchange takes place between phosphate (aqueous phased )an tetraphenylarsonium molybdophosphate (organic phase) in the

presence of molybdate in the aqueous phase. Zeman and Kratzer

[10] applied this isotope exchange reaction to the determination of phosphorus, and Vläc'il [11] applied the method to the

92 determination of trace amounts of phosphorus in platinum

alloys. In this paper, the direct exchange method proposed by

Zeman and Kratzer is modified and made more simple and rapid by

appling the reverse method, and the method is applied to the determinatio f phosphoruo n n somi s e environmental water samples.

3.1 Experimental

3.1.1 Tetraphenylarsonium molybdophosphate solution (TPA-

PMo solution)

10 ml of sulfuric acid (0.1M), 2 ml of ammonium molybdate

(0.02M) and 33 ml of water were added to 2 ml of diammonium

hydrogen phosphate solution (0.02M). The solution was then

shake a separator n i n y funne f tetraphenylarsoniuo ll m wit 0 3 h m

chloride solution (0.001M in 1,2-dichloroethane) for 3 min to

extract phosphorus substoichiometrically in the form of tetra-

phenylarsonium molybdophosphate ((TPA) 3PMoie 2OiTh t0 )• phosphorus concentratio f thio n s organic solution depende th n o s

substoichiometric amount f tetraphenylarsoniuo s m chloride used,

and corresponds to about 10 ygP/ml. 20 ml of the organic phase

was then diluted to 200 ml with 1,2-dichloroethane (TPA-PMo

solutio. A) n

The reagent solution of the lower phosphorus concentration s prepare wa e reagenth f y dilutino db t l 0 m solutio 10 0 1 go t A n

ml with 1,2-dichloroethane (TPA-PMo solutio. B) n

Because partiath f o el solubilit f 1,2-dichloroethano y n i e water, aqueoul al s solutions used were preequilibrate y shakinb d g

with 1,2-dichloroethane.

93 3.1.2 Mock sample solution

Potassium dihydrogen phosphate dried at 110°C was dissolved

ie nsolution th wate d an rf severao s l known phosphorus concentrations (0.0ygP/ml0 1 1~ ) were prepared.

3.1.3 Activity measurement e activitTh s measurewa y d most convenientl y countinb y g

bremsstrahlung fro P witm32 well-typa h e Nal(Tl) scintillation

counter. 3.1.4 Procedure A 5 ml portion of a sample solution is taken and 2 ml of

the sulfuric acid-ammonium molybdate solution (the mixturf o e

the equal volume of 0.1M sulfuric acid and 0.02M ammonium

molybdate solution s addedi ) d thean , n spikef o y addinb dl m 1 g

the definite activity (ao) of carrier-free 32P solution (in the

P(\3~ form) e solutioTh . thes i n n TPAe shaketh -f no witl m 2 h

PMo solution (the amount of P contained: Wj) for l min. The

activity of the organic layer (a2) is measured by pipetting a 1 ml aliquot and counting in a polyethylene counting tube.

e amounTh f phosphoruo t e samplth sf eo containe l m 5 n i d

(Kx) can be readily calculated by Eq. (2). In stea f calculatino d . (2)phosphorue Eq ,th y b g s amount more b en ca convenientl a calibratio yf o obtaine e us y nb d curve,

whic draws i h y treatinb n a serieg f standaro s d samples containing

various known amounts of phosphorus in the same way, and plotting

(ao - a2)/a2 value vs. phosphorus amounts. It should give a streight line with a slope corresponding to Wj.. In this

calibration curve method, it is not necessary to know the

exact value of Wi.

94 3.1.5 Spectrophotometric determination of phosphate by the Japan Industrial Standard fJIS) method [13] A sample solution which contains 0.005 ~ 0.15 mg POi+3~ in it is taken in a 50 ml volumetric flask and diluted to 40 ml with water and 5 ml of ammonium molybdate solution (1.5% in 3M H2SOi») is added. Then, 0.25 ml of tin(II) chloride solution (2% in

1.2M HC1) is added and the solution is diluted to the mark and shaken. After abou mins5 1 t ' standing n aliquoe ,a th f o t solutio optican s takea i n n i ne optica lth celd an ll densitf o y the blue coloration is measured at 700 nm with a spectrophotometer.

After the correction for blank, the phosphate amount in the sample solution can be obtained by using the calibration curve.

2 3. Resul d discussioan t n

Several mock sample solutions containing 0.1 ~ 50 MgP and 0.01 ~ 0.1 MgP of phosphate in each 5 ml were analysed by the procedure describe n 3.1.i d 4 using TPA-PMo , solutioB d an A n respectively. The results are shown in Table I. By this

Tabl I e Determinatio f phosphato n e reversth methoy A b eIE e d

3 3 PO. ' taken P04 " found Error (ygp) (PgP) W

so.o 50.9, 48.2, 49.5 + 1.8, -3.6, -1.0 20.0 19.3 -3.5 10.0 10.4 +4.0 5.0 5.0, 5.21 5. , 0, +4, +2 2.0 2.0 0 1.0 1.0, 1.00 1. , 0, 0, 0 0.50 0.51, 0.52, 0.50 +2, +4, 0 0.25 0.26, 0.24 +4, -4 0.20 0.19 -5 0.10 0.098 -2 0.050 0.048, 0.049 -4, -2 0.020 0.019 -5

95 Table II An example of water analysis by the IEA and spectrophotometric methods.

Concentratio f phosphato n e (vigP/1) Sample by this method by JIS method

Lake Kasumigaura (Ibaraki Pref.) 63 63

River Sakura 99 90 Clbaraki Pref.)

Fig 1 . Sampling sites

96 samplea f o phosphatf l )o m 5 n metho(i e P iondn ug 0.0ca s0 5 2~ be determined withi error% ±5 n . S spectrophotometriJI e Bth y c method othee th rn ,o hand 2 , phosphatf o ) ml 0 determine4 e b ~ ionn n ca s(i P d yg wit 0 5 ~h

2 ~ 10% error [13]. In Table II, an example of the analyses of the lake and river water samples collected in the Tsukuba district (Fig. 1) S spectrophotometri y thimethoA b JI IE sd an d c metho s giveni d , showing that thimethoA IE s d give e resultth s n fairli s y good concordance with those obtaine e conventionath y b d l spectrophoto-

metric method.

methoA IE s suc e ha dh Th advantages comparing wite th h

conventional spectrophotometric metho s followsa d : (1) The IEA method reveals higher sensitivity and higher

accuracy. (2) The analytical operation is much simpler and rapider.

e isotopth s A e exchang) (3 e equilibriu attaines i m d quickly, the small variation in the experimental temperature, acidity of the solution or standing time does no harm. On the other hand,

coloration of molybdenum blue is largely affected with these conditions, and rather unstable.

spectrophotometrie th n I ) (4 c method base molybdenun o d m blue coloration, arsenic(V d iron(DI)an e interferin)ar g ions an a largd e amount halogenidef o s s also interfere with coloration. methodA IE e Ith ,n these o harmn ion o d s. Although vanadium, zirconium, titaniu d tin(IIan m ) interfer methoA eIE wite dth h

[10], the contents of these ions are negligible in environmental water samples.

97 (5) By the aid of an automatic sample changer, the activity

measurement done b en ca sautomaticall seriea r f countinfo yo s g methoe samplesth suitabls , i d So .many-sampler fo e s analyses.

4. Speciative determination of phosphorus

The IEA method can be applied to the speciative determination of phosphate, phosphite and total phosphorus in water samples

accordin e differencth e o isotopt gth n i ee exchange behaviros

f eaco h species.

As is shown on Fig. 2, phosphate ions exchange with TPA-PMo

in the presence of ammonium molybdate in the acid region below

pH 2, while phosphite ions do not. It is also confirmed that

there occurs no isotope exchange reaction between phosphate and phosphite in this pH region. Therefore, when a water sample is analysed by the procedure written in 3.1.4, only phosphate can

be determined even in the presence of phosphite ions. On the other hand, whe e samplth n pretreates i e d with bromino t e

oxidise phosphit phosphateo t e d the,an n treated accordino t g the above-mentioned procedure, the sum of phosphate and phosphite can be obtained.

PO:3- 50f

12345 pH

Fig 2 . Dependenc f isotopi y e exchangH p n o e

98 Whee samplth n e wate s heatei r d wit mixee th h d acid solutio f sulfurio n c acid nitrian d c acid, unexchangable phosphorus including organic phosphoru s decomposei s e th o t d form of exchangable phosphate, and thus total phosphorus can

be determined.

1 4. Experimental

4.1.1 Procedure for the determination of phosphite samplea f o l ,m 5 severa o T l drop f saturateo s d bromine water is added to oxidise phosphite to phosphate. The excess of bromine is eliminated by extracting with 1,2-dichloroethane. The solution is then analysed according to the procedure

writte x valuW 3.1.4n i e n eTh thu. s obtained correspondo t s

the sum of phosphate and phosphite originally present in the

sample. Phosphite contene readilb n ca ty obtainey b d

substracting the phosphate amount which is determined with

another 5ml portion of the sample as described before.

4.1.2 Procedure for the determination of total phosphorus

To 5 ml of a sample, 5 ml of mixed solution of 9M

sulfuric acid and 14M nitric acid (10:1 in volume) is added and

gently boiled for 2 hr to decompose unexchangable phosphorus.

After cooling, the solution is diluted with water *o 100 ml. A

l aliquom 5 takes i td tota an n l phosphoru determines i s d according to the procedure written in 3.1.4.

4.2 Result and discussion

Several mock sample solutions which contains 0.05 ~ 50 ugP of phosphorous acid in each 5 ml were analysed. Results are show Tabln i n. HI e

99 Table HI Determination of phosphite by the IEA method

3 take" PO n P03 " found Error (ugP) (PgP) (*)

50.0 50.8 »1.6 5.0 5.1 + 2 1.0 1.0 0 0.50 0.52 +4 0.25 0.24 -4 0.050 0.048 -4

The mixture solutions of the definite amounts of phosphate

and phosphite were analyse e above-mentioneth n i d d way, giving

e resultth s show n Tabli n . IV Phosphate phosphitd an e b n ca e

determined with in ±5% error.

Table IV Determination of phosphate and phosphite in their mixture

PO/' PO/' PO/" PO/' Error Error taken taken found found (%) (UgP) (UgP) (UgP) (UgP) (*)

1.00 1.00 1.00 0 0.95 -5 1.04 +4 0.97 -3

1.00 0.50 0.96 -4 0.48 -4 0.99 -1 0.52 + 4

1.00 0.25 0.95 -5 0.26 +4 1.03 + 3 0.25 0

e resultso-callee Th th f o s d "addition test" wite th h actual river water sample e showthesn I ar s Tabln i ne . V e tests, the water samples were first analysed,, and then the

known amount f phosphato s d phosphitan e e weree addeth o t d

water samples and analysed. It is revealed that the river

water samples can be analysed within ±5% error suffering from no interference by the foreign ions.

100 Table V The results of the addition test

3 3 O P contents P04 ' added P04 " found 43_ Error Sajnple O P contents PO 3~ added POj3" found (%) (UgP/Sml) (ugP/Sml) (UgP/Sml)

Rv. Sakura 1 0.22 0.10 0.33 +3 0.00 0.10 0.10 0

Rv. Sakura 4 0.24 0.10 0.34 0 0.00 0.10 0.10 0

Rv. Sann2 o 1.08 1.00 2.10 + 1 0.00 1.00 1.02 + 2

Rv. Sann3 o 0.71 0.50 1.16 -4 0.00 0.50 0.51 +2

An example of the speciative determination of phosphorus r phosphatefo , phosphit d totaan e l phosphoru n rivei s r water sample s givei s Tabln i n. VI ePhosphit t detecteno s wa en i d all samples s showtestedi methoA t nI IE tha.f e o d th t

speciative determination of phosphorus offers the useful tool to the environmenta d geochemicaan l l problems.

Tabl I V e Speciative determinatio f phosphoruo n n Rivei s r Sanno (ugP/ml) 3 Sampling Total P 3- point P°4 ».-

1 0.41 ± 0.01 0.01 ± 0.01 N.D. 2 0.5 0.0± 9 2 0.22 * 0.01 N.D. 3 0.5 0.0± 1 1 0.14 ± 0.01 N.D.

4 0.88 ± 0.01 0.46 ± 0.02 N.D.

5 0.66 ± 0.01 0.02 ± 0.01 N.D.

101 . 5 Conclusion

It is shown that the IEA method can be applied to the

speciative determination of phosphorus, and the method has many

advantages comparing with the conventional spectrophotometric metho discusses a d 3.2n i d .

e reversTh principlA IE e e proposed here make methoe th s d

much simple y eliminatinb r preparatioe th g radioacitvf o n e reagent whic necessars i he direc th methodA r IE t fo y . Thust i , is preferable thae inactivth t e sampl s analysei ee th y b d reverse method rather than by the direct one. methoA IE s man e ha dy Th advantages whic e listear h s a d follows:

(1) Specificity or selectivity is high. e procedurTh ) s simpli (2 e d rapid particularn an eI . , when the heterogemeous (two-phases) exchange syste s appliedi m e th , chemical separation of the exchanging components after the

exchange equilibrium is unnecessary. Only necessary is the

physical phase separation. e interferinTh ) (3 g foreign iongeneralle ar s y few.

(4) The speciative determination is possible by using the differenc e isotopth n i ee exchange behavior f eaco s h species.

thers A e ) occur(5 o chemican s l reactio exchangine th n i n g system, the chemically unchanged reagent once used could be used repeatedly after eliminating radioactivity by standing for

radioactive y shakindecab r o y g with e counteexcesth f o sr

species.

102 References

[I] A. Langer, Anal. Chem., 22_ (1950) 1288

[2] N. Suzuki, J. Chem. Soc. Japan, 80_ (1959) 373

[3] S. Hirano, A. Mizuike, E. Nakai, Radioisotopes, ^5 (1964) 118

[4] T.H. Handley, Anal. Chem., ^6 (1964) 153 . IkedaN Amano. ] S ,[5 , Radioisotopes (1967^ 16 ,7 )31 [6] J. Stary, K. Kratzer, A. Zeman, J. Radioanal. Chem., _5 (1970) 71 . KratzerK ] [7 , Radiochem. Radioanal. Letters (1970_ ,_5 9 )6

. ZemanA Star^. ] ,J [8 Kratzer. ,K , ibid. 4 ,(1970 )l

[9] J. Stary", K. Kratzer, A. Zeman, ibid., 6^ (1971) l

[10] A. Zeman, K, Kratzer, ibid., 27_ (1976) 217

[II] F. Vlâcil, J. Raioanal. Chem., £8 (1980) 221

[12] G.A. Perezhogin, L.P. Sidorova, Zh. Anal. Khim., 24. (1969)

570

["13] Japan Industrial Standard, JISK0102 (1981)

DISCUSSION

J. Tölgyessy. What are the disadvantages of the isotope exchange methods?

N. Ikeda. The method is applicable to only a limited number of elements. It still remains to be developed in the future.

103 RESIN BEAD METHODOLOG APPLIES YA O DT FUEL BURN-UP AND FISSILE INVENTORIES*

D.H. SMITH, R.L. WALKER, J.A. CARTER Ridgk Oa e National Laboratory, Oak Ridge, Tennessee, United States of America

Abstract A new technique has been developed that allows acquisition of samples from matrices difficult to access. While the examples given in this paper are from the nuclear field, the technique is readily modifie addreso t d s other areas e techniquTh . e involves obtaining samples on resin beads; each bead then comprises a sampl masr efo s spectrometric analysis. Throug applicatioe hth n of isotope dilution, concentrations of the target elements can be obtaine addition i d theio t n r isotopic compositions. Examplef so application of this technique are given for U, Pu, and Nd.

1. INTRODUCTION Operation of reactors generates situations in which it is very difficul o obtait analytican a n e radioactivitth l samplo t e edu y of the matrix. Such situations arise, for example, in the nuclear fuel cycle, where it is important to obtain samples from all stages to establish a material balance for fissile components, and in evaluating reactor performance by determining burn-up. The conventional analytical approach involves time-consumind an g costly elemental separation o obtait s n uraniu d plutoniuan m m individually. chemistre th Mos f o t y involved mus e carrieb t t ou d e intensth o t et cel e ho radioactivit ildu a n e samplth ef o y matrix. It is highly desirable, therefore, to develop a technique that both simplifies and expedites sample acquisition. o majoTw r areas benefit from suc techniquea h : various safeguards programs chargee thaar t d with monitoring fissile componentn i s the nuclear fuel cycle, and programs that monitor reactor performance. Safeguards programs must follo course th w f uraniueo d an m plutonium through the fuel cycle from mining to reprocessing; it is at reprocessing plants that the new technique has had its greatest impact. Reactor operation is monitored by measuring the concentratio a fissio f o n n product e hav;w e used neodymiur fo m this purpose e samplTh . e matri s fuei x l thas beeha tn burnea n i d reactos thui sd highlan r y radioactive. We have develope a techniqud e involving adsorptio f targeno t element n anioo s n resin beads.* Each bead comprise a ssampl r fo e isotopic analysis via mass spectrometry. Although any amount up to a microgram or two of analyte can be adsorbed on each bead, most of our work has been in the range of 1-3 nanograms. This quantity of material reduces radiation hazards and greatly simplifies shipping since the quantities involved do not require the costly shielding otherwise necessary.

Researc* h DepartmenS sponsoreU e th y db f Energy o t , Offic f Basio e c Energy Science Officd an sf o e Safeguards and Security, under contract DE-AC05-84OR21400 with Martin Marietta Energy Systems, Inc.

105 > ' . MAS2 . S SPECTROMETRY Because the amount of sample is small, a highly sensitive mass spectromete s i requiredre timth were ew t A e. first workinn o g developing the technique, specialized instruments were required, such as the ORNL multi-stage mass spectrometers.2 These instruments are equipped with pulse-counting detection systems that count each ion individually; one nanogram of uranium is adequate analysisn sampla r fo e . They also have high abundance sensitivity to allow measuremen f o isotopit c abundance e part-perth n i s - million range. Another requiremen s higi t h sample throughput, since the areas to be monitored generate large numbers of samples. Thi usualls i s y addresse mountiny b d numbega f samplo r e filaments on a carousel and mounting it in the ion source chamber of the mass spectrometer.3 This provides the capability of analyzing each sample on the carousel without breaking vacuum; capacities of these carousels range from six to sixteen. An alternative sampla e approacus e o t insertio s i h n probe that allows introduction of the sample through a vacuum lock, but this system is extremely difficuls beeha n e limited us o automatt s it d . an e r originaSinc ou e tim f th eo e l work (1974), commercial instruments have appeare markee th n to d that have sensitivityth e , both with regard to sample size and to abundance, to perform these analyses; they operate under computer control and have throughput superior e specializedtoth thaf o t , custom-built instruments originally used. VG Isotopes and Finnigan MAT both market instruments with these properties.

3. DATA REDUCTION The calculations required for analyses of the sort under consideration here demand automatic execution by a computer. Some of the operations are logical]- complex, and the speed of a computer is necessary to avoid having manpower tied up in routine manip- ulation of data. Extensive FORTRAN programs have been written to allow automatic processing of data and production of a report of results.operatione th l Al 4 s described belo e incorporatear w n i d these programs. 3.1 Uraniu d plutoniuan m m Processing data from these two elements requires a correction to mas8 23 bse positioe applieth o t subtraco dnt e contributioth t n of the contaminant element from the analyte element. Thus uranium and plutonium from resin beads are treated in pairs, one analysis for each element. Approximate values for the mass 238 position are obtained wit correctiono n h s applied. Then, usin e 235/23th g 8 ratio from the uranium analysis as a basis, 238^ j_ subtracted

from the plutonium data, and the plutonium data reprocesses d to arrive at a composition for the final report. The corrected 238/239 ratio for plutonium is then used to correct 238U position for plutonium uraniue th ; m date reprocessear a a fina d lan d report generated usin correctee th g d values. 3.2 Neodymium Neodymiu s interferenceha m s from cerium. Samarium also caused problems until a satisfactory separations scheme was developed; it is described below. mas4 Ceriu14 s d an m 2 contribute14 e th o t s positions of neodymium. Mass 142 is used to subtract the contrib-

106 ution of natural neodymium (a contaminant in this case) from the fission product isotopes. Mass 144 is used only if a total neodymium concentratio s requiredi n prograe Th . m first processes the uncorrected neodymium datd thean an processe ceriume th s . After the cerium results are available, correction is applied for that element to the neodymium results obtained earlier, and the natural neodymium correctio s i napplied A repor. t e sheeth f o t corrected composition is produced, and burn-up calculations are performed accordin standara o t g d ASTM protocol.^ 3.3 Isotope dilution Isotope dilutio widels i n y use masn i d s spectrometr determino t y e the concentration of the analyte element. A known amount of an isotopically enriched isotop s knowa addei e o nt d e amounth f o t sample. Knowledge of the isotopic compositions of the spike, the unspiked e mixturth sample d f spik o an ed , sample an e , together with the weights of spike and sample, allow calculation of the concentration e prograTh . m thus requires e isotopiaccesth o t s c composition f spiko s d samplan e e whe mixtura n s beini e g analyzed. When this requirement is added to that of analyzing mixed beads of uranium and plutonium, a program of some complexity is required.

We most often use spikes of 233 242pj 150^ j. f U anc u Nc 3.4 Internal calibration Using a double spike for internal calibration improves precision by a factor of five or more.^ In this technique, a spike containing two enriched isotopes whose ratio is known to high degree of accuracy and precision is added to the sample. By comparing the measured value of this calibration ratio to the known, a bias can e calculateb d tha s applicabli t e analysith o t equestionn i s n A . average bias is applied if internal calibration is not used; experienc s showha e n that suc averagn a h e doet usuallno s y reflect the conditions of the individual analysis. The resin bead techniqu s fulli e y compatible with this technique l thaal s ;i t necessary is to implement the necessary calculations in the software.

4. SAMPLE ACQUISITION 4.1 Uraniu d plutoniuan m m Spent reactor fuel is highly radioactive, being comprised of fission products, transuranium elements (including plutonium)d ,an unburned uranium fuel. Because importancth f o e f safeguardo e s programs and because of the number of samples that need to be analyzed e resith , n bea ds greatesit techniqu d ha s t eha impac n i t this area. To obtain samples of uranium and plutonium from solutions, the fact that these elements form stable anions in nitric acid solution is exploited. Figure 1 shows the adsorption curves of various elements of interest from 8M nitric acid.''' The elements that adsorb appreciably are U, Pu, Th, and Np. Most fission products and actinides higher than neptunium remain in solution. Neither neptunium nor thorium interferes with subsequent mass spectrometric analysis: neither has an isotope that overlaps one from uranium or plutonium, and neither is radioactive enough to cause problems on that score. Indeed, if a thorium-based fuel

107 u ' NO ADS NO ADS 8 S\ / "N ^j^ ••»^

4 Ëto Th u NP/ Pu, Am Cm 0 tf \ ff \ 3 S / / X m DE § 2 / ni NO/ kDS NO ADS 1 / «•N ! / / ^ f 0.11 5 10 14 on 5 10 14 an 5 1014 an 5 10 14 an 5 1014

HN03, M IONIC METAL ADSORPTIO STRONGLY B N Y BASIC ANION EXCHANGE RESIN ( Dowex 1-X2 )

FIG 1. Curves for adsorption from 8M HNC>3 of various elements on anion resin beads.

cycle is ever implemented, it will be advantageou havo t s e that element alreade beadth .n yo To obtain samples, an aliquot of sample solution is adjusted to 8M in nitric acid A numbe .f resi o r n bead s i sintroduce e th o t d solution; the number will depend on the application. For operations at reprocessing r_Jnts, where large quantities of the analyte elements can be expected to be in the environment, we have modified our original technique to accommodate 1000 beads by samplin a solutiog n containin f uranium.o g m 1 g ^ Even thougt i h is difficul o envisiot t scenaria n o where 1000 samples woule b d needed, this greatly reduce e effecth s f o contaminationt e Th . solution-bead mixtur s i agitate evortea n o xd mixe r 10-2fo r 0 minutes; this is long enough to allow adsorption of 1-3 nanograms eac f o huraniu d plutoniuan m n o meac h bead e bead Th e the. ar sn separated from the solution. The level of activity on the beads is low enough that they can now be removed from the hot cell. This wor f specializeexpedites o i k e us y b d d equipmen e havtw d ha e constructe r specificationsou o t d . Figur s photograpi 2 e f thio h s equipment e bead Th e caugh .filtea ar s n o t r inserte e smalth n li d funnel; this apparatu s removei s t cel dd ho seale lan froe th mt a d each end with plugs. It is then a convenient package in which to shisamplese th p . Because such small quantitie f uraniuo s d an m plutonium are involved, no special shielding is required. It is lega shio t l p such sample e numbemaiy sb th s lon f beada lo s a rg s is not too great; this results in great savings in safeguards programs where several thousand samples are analyzed annually. 4.2 Neodymium Unlike uranium and plutonium, analysis of neodymium for reactor burn-up measurements is usually done at a laboratory near the site reactore oth f . Other element usee r b thi fo dn sca purposee ,th

108 FIG.2. Resin bead separation d shippinsan g apparatus.

109 choice being dictatee reactornature th th f o e y e havb d;w e used zirconium to monitor burn-up in operation of the ORNL high temperature gas-cooled reactor.^ majoe th rf o difficultie e On usinn i s g neodymiu y othean r o m rare earth element for this purpose is the presence of isobaric interferences r neodymiumfo ; , these arise froe presencth m f eo samariu samplee ceriud th an m n i m: they, too e fissioar , n products producee an e reactoar d th n i dr simultaneously with neodymiumy B . extracting neodymiu n anioo m n resin bead y agitatinb s g them wita h methanol-nitric acid medium (90% methanol, 10% 1:1 HNOß-water), separation from fission products heavier than neodymium is achieved. This eliminates samarium froe sampleth m e ; th thi s i s most troublesome contaminant becaus s ionizatioit e n characteristics are similar to neodymium. Cerium still remains and must be corrected for in the data-processing step; this requires measuring the isotopic composition of cerium at the conclusion of the neodymium measurements.

5 LOADING SAMPLE N FILAMENTSO S As mentioned previously, each resin bead constitute sa sampl r fo e mass spectrometric analysis have W . e foun optimue th d m bead size to be in the range of 150-250 urn. Beads of this size are diffi- e wit unaidee se cul th ho t d e perforeyew o s , m bead manipulations under microscopes. Single beads can be identified and picked up on sharp needles; these are made of tungsten or stainless steel to prevent chemical attac y nitrikb c acid adherin e beade Th th .o t g bead is then transferred to a mass spectrometer filament; filament e canoe-shapedsar d touchin desiree ,an th e bea o th gdt d spot on the filament effects the transfer. The filament is then crimpe o helt d p hole bean th dplacei d , althougt h no this i s strictly necessary.

6. ISOTOPIC ANALYSIS 6.1 Uraniu d plutoniuan m m We have developed a technique of analyzing plutonium and uranium sequentially froe samth me filament. This involves applyina g correction to the 238 mass position for each element, which of course degrades somewhat the result for this isotope. Because 238 e majoth is r isotop r uraniumfo e , this correctio s littlha n e effect on the quality of the result for that isotope. The uranium correction is often significant for the 238p^ however, since this

isotop s i ealmos t alway minoa s r constituen f to plutonium f I . u results of highest accuracy and precision are needed for 2^°Pu, it is necessary to separate the two elements by washing the beads with water or dilute HNO-j to remove the uranium. Mass 235 is monitored during plutonium analysimonitores i d 9 masan 23 ss d during uranium to provide reference isotopes to use in correcting mas8 th23 es position. Sinc t i ionizea e lowe t a rs temperature (1450-1550°C) than uranium, plutoniu s analyzei m d first. Care mus e exerciseb t o t d avoid raising the temperature too fast; doing so apparently causes the beao t explodd e with some violenc d oftean e n resultn i s excessive los f sampleo s . With reasonable caution, however, this problem is eliminated: a rise in pressure signals the onset of bead disintegration d carefullan , y raisin e temperaturth g n i e small increments through this l tha regios needeal i ts i o nt d

110 assure sample integrity. To avoid excessive uranium' contamination, the count rate for Pu+ is kept in the neighborhood of 100,000 counts per second; temperatures above 1600°C are to be avoided. e problemth f o s e encountereOn e isotopith n i dc analysif so

plutonium is the accumulation of 241^ thfe decay of 241p m u

(half-lif f 14.o e 3 years). This cause difficulto sn beade th sf i y ro e fresar h since complete separatio plutoniuf no m from americius i m achieved in the sample-taking process. If, however, the beads are several months old, significant amount f 241o s ^w ill have accumulated s thei t nI necessar. o burt yamericiuf nof m untila stable mass 241 signal is achieved. This is determined by monitoring the ratio of 241 to a convenient plutonium isotope. Since americium ionizes more readily than plutonium, this problem is usually dealt with fairly easily. When enough data have been accumulated for plutonium, the remain- der of this element is burned off by raising the filament temperatur o t aboue t 1700°C. This usually takes abou minutesn te t t ,bu can requir o heavilto ee s beath longeywa df i loader d with plutonium. Excessive burn-of avoidee fb o timet possif i de sar - ble; we have found that, since 238 is a minor constituent of plutoniu majoe r uraniumth fo r d an m, data-takin uraniur fo g n ca m commence when the 239pu/238y ratio reaches the range of 1-10; the valu n thii e s criterioa usee rang b s a do t efunctioa e s i nth f o n isotopic compositions of uranium and plutonium. Uranium data are accumulated at a count rate of about 300,000 counts per second on the major isotope. s desirablei t Ii f , e uraniuseparateb n ca m d from plutoniuy b m washing the beads with dilute nitric acid; each element can be analyzed individually. 6.2 Neodymium Because burn-up calculations require that natural neodymium be subtracted e fissiofroth m n product component n a accurat, e measurement of l^Nd must be made. Since cerium also has an isotope of this nominal mass, its relative abundance at this mass position must be determined so that the appropriate correction can be made. e samTh e precautions take n o initian l heatin r uraniufo g d an m plutonium mus e observetb r neodymiufo d m analyses A smal. l signal from samariu mast usualls a i m s 0 15 positiony d observedan 8 14 s ; this must be burned away before data accumulation for neodymium can be started. Neodymium runs at about 1500°C; like most rare earth elements s i extremel t i , y well behaved under thermal ionization conditions. When enough data for neodymium have been collected e temperatur,th e excesss i raiseth e o f burt .dof n Cerium is run at about 1700°C; a complete isotopic composition is not necessary since only the 142 and 144 positions are of concern; only masses 140, 142, and, if total neodymium is required, 144 are monitored. Mass 140 is monitored during neodymium analysis to provide a reference for cerium corrections.

Ill 3 lonizatio6. n from resin beads e developmenth Earl n i y f o thit s technique e observe,w d that performance characteristics during isotopic analysis were improved when samples were loaded on resin beads in comparison to samples loaded as solutions. An order of magnitude more sensitivity in collection efficiency was observed. To investigate this phenome- secondara non, n masio y s spectrometric analysi f sresio n beads was performed.1" We found that, during analysis, uranium stayed in the carbon skeleton of the bead and did not migrate to any appreciable extent int e filamente rheniuoth th e bea f Th o md . thus serve a poin s a d t sourc ionsf o e , enhancin collectioe th g n efficiency of the ion source, whose lensing action is designed to operate on a point source.

FIG.3. Electron micrographs of resin beads before and after heating.

112 A rhenium-carbon mixture has a higher work function than pure polycrystalline rhenium.^ lonization occurs with about a factor of ten higher efficiency from such a surface. The interface between the bead and the rhenium filament forms a composite of this nature, and thus ionization will take place largely in this region e foundW .n additioni , , thae carboth t n skeletoe th f no bead slowly dissolves into rhenium substrate; rheniu d carboan m n do not form a stoichiometric compound but are mutually soluble. Figure 3 contains three electron micrographs showing a pristine bead and two beads heated for 30 minutes at 1700°C. Dissolution of the beads from the side in contact with the rhenium is readily apparent; this confirmeswa y elementab d l distributions determined by secondary ion mass spectrometry. Thus the carbon matrix serves aa sreservoi f o rsample , deliverine ionizatioth o t t i g n region as it dissolves in the rhenium. This allows for more controlled evaporatio e sampleth n additionaA f o n. l benefi s gainei t d from e reducinth g é carbonaturth f neo matrix. With solution loadings, los f o ssampl s a oxide e, UOspecieg. . +,(e sn everUO^^ a s i -) present problem. Using resin beads virtually eliminates evaporatio s oxidesa n , contributin o enhancet g d ionization efficiency. lonization efficience b furthe n ca r y enhance y b judicioud s overcoatin e sampleth f o g. While colloidal graphite (aquadags ha ) the desired effect, greater benefit can be derived by overcoating with rhenium, either by electrolytic deposition12 or by depositing a finel s ia t y divided powde slurra s a rn sucros i y e solution.^3 7. RESULTS A study of the comparative efficiency of plutonium ionization is summarized in Table I. As can be seen, efficiencies of about 1% can be realized, making possible acquisition of quality results from analysi f subnanograo s m samples. e Becausdangerth f f o eo s plutonium to human health, it is desirable to minimize the quantity of sample necessary for analysis. Table II presents results for all three elements under consideration here to give a unified example of the technique's performance.

TABL . I EComparativ e lonization Efficiencr fo y Pure Pu Standards Filament loading Amount Lab Technique Loaded, ng % Efficiency ORNL Solution 1.0 0.01-0.05 ORNL Resin bead 1.0 0.10-0.20 ORNL RB/Re-C overcoat 1.0 0.5-1.5 ORNL Sol/Re-C overcoat 1.0 0.02-0.05 LANL Re electrodeposit 10. 0.014 LLNL Solution 0.000125 0.05 KAPL Solution 1.0 0.04

ORNL: Oak Ridge national Laboratory LANL s AlamoLo : s National Laboratory LLNL: Lawrence Livermore National Laboratory KAPL: Knolls Atomic Power Laboratory

13 TABL . IsotopiII E c Compositions froa m Single Reactor Run (Atom %)

Starting Fuel 5 23 ____234 ____8 23 6 23 0.360 45.20 0.430 54.01

Spent Fuel 142 143 144 145 146 148 150 Nd 1.27 8.32 42.43 14.86 19.32 9.055 4.754

238 239 240 241 242 244 Pu 13.85 42.47 18.33 10.60 14.71 0.0023

234 235 236 238 0.296 1.578 11.128 86.998

TABLE III. Result f Joino s t IAEA-PNC-ORNL Resin Bead Experiment

Concentration n ug/i s g Lab Technique Tota U l Totau P l PNC Conventional 164.7 1.507 IAEA Conventional 164.8 1.492 IAEA Resj.n bead 164.7 1.498 ORNL Resin bead 164.0 1.492

Isotopic Composition, weight % U 8 23 6 23 5 23 b La 4 Techniqu23 e PNC Conventional 0.0191 1.0942 0.3746 98.512 IAEA Conventional 0.0211 1.0876 0.3718 98.519 IAEA Resin bead 0.0201 1.0905 0.3699 98.519 ORNL Resin bead 0.0201 1.0905 0.3722 98.511

Isotopic Composition, weight % Pu 9 23 Lab 8 23 Tech 240 241 242 PNC Conv 1.458 60.178 22.623 11.515 4.227 IAEA Conv 1.369 60.301 22.642 11.483 4.206 IAEA Bead 1.375 60.171 22.599 11.624 4.231 ORNL Bead 1.366 60.232 22.635 11.529 4.239

e requirementw th techniqune f o a e s On i f thayielt o ei s t d s tryini t o i replacet g e resulton t e .a leasth ss a goots a d Table III summarizes results from an international experiment involving the International Atomic Energy Agency, the Power

Reactor and Nuclear Fuel Development Corporation of Tokai, Japan, and ORNL.^1 Resin bead result comparable sar thoso et e obtained from solutions C interactioPN ; IAEd an A s continueha n d wita h

series of similar experiments. A resin-bead sample round robin

5 . NationaadministereS . U e lth y Bureab d f o uStandards yielde1 d results that were, on the average, better than those obtained for IDA-80 n internationaa , l round robin using solution loadings.1°

114 CONCLUSION. 8 S We have had more than ten years' experience now using resin bead sample loading. The benefits are so great, and the implementation so simple, that it is now the method of choice for introducing any sample into our pulse-counting mass spectrometers. If the analyte element amenablno s i to separatio t e y resib n n bead methodology, we introduce a cation resin bead to the filament to absorb the solution containing the analyte. In this way, we can exploit the benefits accruing from thermal ionization from resin beads without havin o perfort g y separativan m e chemistry with theme W . know from experience that the operations involving the use microscopes ten o intimidatt d e prospective users., is Theie us r however, much less formidable than it appears: any technician can learn the necessary procedures in a few days, and, with experience, introducing sample o t filaments n resio s n beads becomes quicker than as solutions, taking only a few seconds. We have usee techniquth d n i mane y applications other than those describe n thii d s paper. These include e Thresupporth er Milfo t e Island incident- d determinatioan ^ f technetiuo n e environth n i m- ment. 18 & comparative study of the uptake of uranium, thorium, and plutoniu y plantb md animal an s bee ha sn published.-e Th 9 three elements were analyzed sequentially from the same filament; thorium was analyzed after uranium. The range of applications of the resin-bead/nass spectrcmetry combination covers a wide variety of elements and sample matrices. It provides a convenient shipping medium as well as a handy vehicle for introduction of samples into the mass spectrometer. Other major advantage e techniquth f e o ssimplifiear e d sample preparation, both with regard to time and chemical operations; improved mass spectrometric performanc o enhancet e du e d ionj zation efficiency; reductio contaminanf o n t species suc 241s o a ht e ^u ^ the separative powers of the beads; reduction of loss of sample as volatile oxide n bote abilitspeciesru th h d uraniuo t yan ;d an m plutonium froe samth me filament. Thes e powerfuar e l arguments for widespread adoption of the technique.

REFERENCES 1. Smith, D. H., Walker, R.L., and Carter, J. A., Anal. Chem. 54 (1982) 827A. 2. Smith, D. H., Christie, W. H., McKown, H. S., Walker, R. L., Hertel, IntR. . Mas. J G , s Spectrom n Phys Io 0 (1972-731 .d an . ) 343. 3. Christie Cameron, . , RevH E. . ,W . . A , Sei. Instrum 7 (19663 . ) 336. 4 ., Chem.. SmithH . D ,, Biomed d Environan ., . Instrun-. 0 (19801 . ) 27; USDOE Report ORNL/TM-8356, August, 1982. 5. ASTM procedur E e321 , American Societ r Testinfo y Materid an g - als, Philadelphia, PA, revised annually.

15 6. Institut. J Walker, Smith, H. L. . D . f Nuclea,R o e r Materials Management 13 l (1985) 44. 7. Faris, BuchananP. Anal, . ,J F. . , R Chem 6 (19643 . ) 1158. 8 .d Smith, Anal Walker an H. , Carter , . L. , D A. . Lett,R . ,J 4 1 . A19 (1982) 1603. 9. Walker, R.L., Botts, J. L., Carter, J. A., and Costanzo, D. A., Anal. Lett4 (1977 0 1 . ) 251. 10. Smith, D. H., Christie, W. H., and Eby, R. E., Int. J. Mass Spectrom. and Ion Phys. 36 (1980) 301. 11. ChemJ . , Smith.H. Phys. D ,5 (1971 5 . ) 1482. 12. Rokop, D. J., Perrin, R. E., Knobeloch, G. W., Armijo, V. M., and Shields, W. R., Anal. Chem. 54 (1982) 957. 13d Carter ., Intan SmithA. . . Mas.H J . ,J . s,D Spectromn Io d an . Phys 0 (19814 . ) 211. 14. Walker, R. L., Smith, D. H., Carter, J. A., Donohue, D. L., Deron , AsakuraS. , , Kagami,Y. , Irinouchi d MasuiK. , an , J. , S. , J. of the Institute of Nuclear Materials Management 10 3 (1981) 43. 15 .d KellyAnal, an FassettR. , . ,W D. Chem . 6 ,J (19845 . ) 550. . Beyrich, Golly16 A. , Spannagel W. . ,W , , DeBievreG. , , P. , Wolters, W., Gallet, M., Machlan, L. A., Grämlich, J. W., and Fassett, J. D., in Proceedings of International Symposium on Recent Advances in Nuclear Materials Safeguards, Vienna, Austria, November, 1982: IAEA/SM 260/33. d Christiean . CarterWalker, , 17 , SmithH. A. . L. W , . . . D J ,,R H., Int. J. Environ. Anal. Chem. 8 (1980) 241. 18. Anderson, T. J., and Walker, R. L., Anal. Chem. 52 (1980) 709. . Garten , Bondietti19 T. . C , , . E.A.EnvironJ Walkerd , an ,L. . R , Quality 10 (9181) 207.

DISCUSSION

T. Braun. I would be interested in the type of resin beads you were using. Were they microreticular or macroreticular? I think swelling could have a certain role in the disintegration of the beads during the heating before the MS analysis.

D.H. Smith. Thee macroreticulaar y r beads (125-250 micrometer n diameter)i s . e sizs choseTh wa e o makt n e loadin gn quantitie i bead u P sd witan s U h reproducibly variable fro 1 nanogram l microgram o t m .

116 A SURVEY OF IMMUNOASSAY TECHNIQUES FOR BIOLOGICAL ANALYSIS

C.A. BURTIS Oak Ridge National Laboratory, Oak Ridge, Tennessee, United State f Americso a

Abstract

Immunoassay is a very specific, sensitive, and widely applicable analytical technique. Recent advance n genetii s c engineering have led to the development of monoclonal antibodies which further improve e specificitth s f immunoassayso y .

Originally, radioisotopes were use o labe t de antigen th l d an s antibodies use n immunoassaysi d . However e lasth t n i ,decade , numerous types of immunoassays have been developed which utilize enzyme d fluorescenan s t dye s labelsa s . Give e techth n - nical, safety, health, and disposal problems associated with using radioisotopes, immunoassays that utilize the enzyme and fluorescent label e rapidlar s y replacing those using radioisotope labels. These newer technique s sensitivea e ar s e ar , easily automated, have stable t havreagentsno a eo d d ,an disposal problem.

In 1959, Berson and Yalow [1] reported the development of a revolutionary new type of assay which combined the specificity inheren e antibody/antigeth n i t n interactio e excellenth d an n t sensitivity obtainable from radioisotope measurements. Since that time, immunoassays have become one of the most important and widely used of all analytical techniques. A recent study by Frost and Sullivan [2] reported that in 1983 the U.S. market

117 for immunodiagnostic reagents and instruments exceeded

$600 millio d $10an n 5 million, respectively.

The success of immunoassay as an analytical procedure is

due to the exquisite specificity and the high degree of affinity

of the antigen/antibody interaction. Thus, the antibody, or

e antiserth s n y whiccontainedreageni i ake n a t i e h f th o t s i ,

immunoassay. Initially, antibodies were obtained by injecting

animals, usually goats or rahbits, with a quantity of purified

r semipurifieo d immunogen. Afte a giver n time interval, anti-

bodies coul e harvesteb d y processinb d g blood taken froe th m

immunized animals. The resultant serum was then used directly

n assaa r furthen o i y r purifie y varioub d s precipitation,

adsorption r Chromatographio , e techniques. Since antibodies

e biologicaar l compounds, their variability, stabilityd an ,

availability have bee f concerno n a grea d tan , dea f efforo l t

s beeha n expende n ensurino d d documentinan g e essentiath g l

propertie d characteristican s f antibodo s y reagents (i.e.,

affinity, avidity, specificity, stability, etc.).

In 1975, Kohler and Milstein [3] reported their develop-

ment of a cell fusion technique that has revolutionized the

immunodiagnostic industry. In this new technology, two dif-

ferent type f cello se fuse ar s o creat w t typede ne Th a e.

resulting cell, calle a hybridomad , exhibits characteristics

derived from both parent cells r exampleFo . , fusing fast-

growing myeloma tumor cells with spleen cells derived fron a m

animal that has been immunized with a specific antigen yields

hybridomas that secrete antibodies against this antigen.

Subsequent selection from the population of fused cells can

isolate hybridomas that secrete a single type of antibody with

118 a unique set of biochemical characteristics and properties.

Sinc e antibodth e a specifi s i y c immunoglobin fro n immunoa m -

globin gene of a single cell, it is called a "monoclonal

antibody." In addition to antibody-producing capabilities, the

hybrodomas acquir e growtth e h characteristic e myelomth f o sa

e propagateb n parenca d y celb dan t l cultur r otheo e r techniques

o providt e continuous quantitie e monoclonath f o s l antibody.

As immunochemical reagents, monoclonal antibodies havo tw e

significant advantages over conventional antibody reagents:

(1) they have a unique and narrow antigenic specificity, and

e use b o produc) thet dn (2 ca y e large quantitie f essentiallo s y

identical antibodies. s Thuspossibli t i , o product e e large

quantitie f monoclonao s l antibodies, maintaining both con-

sistent yield d lot-to-loan s t quality.

UNKNOWN SPECIFIC ANTIGEN rANTIBODY

RADIOACTIVE ANTIGEN

PERCENTAGF EO RADIOACTIVITY BOUND TO ANTIBODY

CONCENTRATION OF ANTIGEN

FIG. l. RADIOIMMUNOASSAY (COMPETITIVE PROTEIN BINDING)

The theoretical n iinmunoassabasia f o s s i illustratey n i d e followinth y b d gan FIG 1 formula. :

Ag + Ag* + Ab t AgAb + Ag*Ab , (1)

119 where

Aunlabele= g d antigen (i.e., analyte);

= labele * Ag d antigen;

A= antibodyb ;

AgAb = unlabeled antigen/antibody complex; and

Ag*A = blabele d antigen/antibody complex. e absencth n f unlabeleI o e d antigen (Ag) quantita , f labeleo y d

antigen (Ag*) will bind to the antibody (Ab) according to the

law of mass action. In the presence of unlabeled antigen, the

amoun f antibody-bouno t d labeled antigen (Ag*-Ab) will decrease

because the unlabeled antigen competes with it for the binding

sites on the antibody. The amount of unlabeled antigen

displaced froe antibodth m s proportionai y e amounth f o o t tl

unlabeled antigen in the system. Using known quantities of

unlabeled antigen a calibratio, ne prepare b curv n d ca usee an dd

to quantitate the amount of unknown antigen contained in a

sample.

Various labels are used in immunoassays (TABLE I). Radio-

isotopes were the first labels used in immunoassays [1] since

the capability then existed for measuring very low levels of

radioactivity. Isotopes used include 125I, d 131an I, H ,3

occasionally ll+C. As a consequence of combining the specifi-

TABL . I ELabel s usen i d immunoassays ( IA)

1. Radioisotope (RIA)

2. Enzymes (EIA)

. 3 Fluorescent dyes (FIA)

. 4 Chemiluminescent compounds

5. Stable free radicals

120 e antibody/antigeth cit f o y n bonding wit e sensitivitth h f o y

radioactive measurements, radioimmunoassay (RIA) quickly became

a very popular bioanalytical technique and widely used for

determining hormones, vitamins, drugs, cancer antigens, enzymes,

receptors, viruses, antibodies, polypeptides, and other proteins.

Unfortunately, RIA has several inherent problems — all

associated with the use of a radioactive label. First, the

specific activit e labeleth f o yd antigen will decay with time,

and, consequently, the stability and reliability of the key

analytical reagene procesth n s i continuallti s y changing.

Thus e shel,th e freagen th lif f s o shortei t . Secondly,

RIA-type assays have been traditionally difficul o automatt t e

because the technique requires a separation step of the free

and bound portions of the labeled antigen before they are

counted. Several attempts have been mad o automatt e A RI e

methods, but, in general, these efforts have not been as

successfu e developmenth s a l f automateo t d equipmenr fo t

routine chemical assays.

A third disadvantage of RIA is the safety and health

concern associated with the use of radioactive compounds. Such e requireus s licensin d regulatioan g e laboratorth f o n y b y

various state and federal agencies including the Nuclear

Regulatory Commission, the Department of Transportation, and

the Food Druan dg Administration n additionI . , bote th h

laborator d technologistan y s shoul e continuallb d v monitored

usin a combinatiog f areo n a monitor d personaan s l area moni- alsis tors oIt becomin. a widespreag d laboratory practicto e

prohibit pregnant women from workin n laboratori g y arean i s

which radioisotopes are used.

121 An additional disadvantage of RIA, which is rapidly

becoming its MAJOR DISADVANTAGE, is the disposal of its

radioactive waste r manFo y. years, such wast s disposeewa f o d

by a wide variety of means, including pouring down the drain.

However, in recent years, disposal of radioactive waste has

been strictly regulated, which has led laboratories to consider

alternative analytical techniques.

n 1972I , Rubenstein, Schneider d oilma an ,reporte] [A n a d

new type of immunoassay in which the radioactive label was

replaced by an enzyme. Such an assay is called an enzyme

immunoassay (EIA) and is classified as a homogeneous assay

becaus t doei t erequir no s e separatioth e f freo d nboun an e d

labelA techniquRI s doee a s th s e (i.e. ,heterogeneoua s assay).

The theoretical basis of an EIA is illustrated by the

following formulas:

Ag + Ag-K + Ah t AbAg + AbAg-F , (2)

AbAg-E + S t AbAg-E + P , (3)

where

Ag = unlabeled antigen,

Ag-= enzymE e labeled antigen,

AbA = unlabeleg d antigen/antibody complex,

AbAg- = enzymE e labeled antigen/antibody complex,

= enzymS e substrated an ,

P = product of enzyme-catalyzed reaction of the substrate.

n practiceI n aliquoa , f specimeo t s adde i nd mixe an d d

with aliquots of enzyme-labeled antigen, antibody, and the

substrate appropriate for the enzyme. If there is no unlabeled

122 antigen in the specimen (i.e., the analyte being measured), the

enzyme-labeled antige d antibodan n y react, formin a complexg

that blocks the active center of the enzyme and thus inhibiting y enzyman e reaction e unlabeleth f I . d antige s preseni n n i t e specimenth wilt i , l displac e enzyme-labeleth e d antigen i n

proportio s concentrationit o t n e displaceTh . d enzyme-labeled

materia s thei l n fre o bint e d with substrate, permittine th g

enzyme reaction to occur. The resultant enzyme activity can be

related to the concentration of the unlabeled antigen by

comparison with a calibration curve.

Sensitivity of an EIA can be increased by allowing the

enzym o react e t wite substratth h a longe r fo er perio f timeo d .

Thus, the product increases in a stoichiometric manner, and the

signal-to-noise ratio (and hence, the sensitivity of the assay)

increase accordingly. Such assays are known by the acronym

EMIT, which stand r Enzymfo s e Multiplied Immuno Technique.

Another version of EIA is the enzyme-linked immunosorbent

assay (ELTSA), which is widely used to detect the presence of

either antigen r antibodieo s n biologicai s l samples. This

method utilizes enzyme-labeled immunoreactants (antigenr o

antibody) and a solid-phase binding support (e.g., coated test

tubes, bead r wello s f microtiteo s r plates) n variouI . s types

of assay modifications, the amount of immobilized enzyme

activity is proportional to the concentration of the analyte

being measured. Nano d picograan - m quantitie f analyto s n ca e e detectedb , making ELIS s sensitiva A s RIAa e .

Fluorescent labels are also becoming increasingly used in

immunoassays. Fluorescent immunoassays (FIA) utilize

fluorescent labeled antibodies or antigens to detect antigen-

123 antibody reactions. Measurement of the resultant fluorescence

can be used to quantitate the concentration or titer of either

the antige r antibodyo n A variet. f technicao y l variations

exist for FIA (TABLE II), with fluorescence polarization being mose th t f usefuo e on l [5].

TABLE II. Types of fluorescent immunoassays*

Fluorescence microscopy

Heterogeneous:

Solid-phase antigen assay

Solid-phase antibody assay

Homogeneous:

Antibody quenching

Fluorescence polarization

Internal reflection spectroscopy

Fluorescence excitation transfer

Fluorescence protection

*From Dictionary and Encyclopedia of Laboratory Medisi-ne and Tech- nology, . BenningtonL . ,1 , Ed., W. B. Saunders, New York (1984) 785.

Assays utilizing fluorescence polarizatio e basear nn o d

the fact that light waves produced by standard excitation

source e orientear s d randomly. Also, light waves passed

through certain crystalline substances (polarizers) become

oriente a singl n i de plane-polarized b e sai o ar plant d d an e .

If a fluorophor is excited by polarized light, the size of the

molecule will determine if the fluorophor emits polarized or

depolarized light. If the fluorophor is large, the rotational

relaxation time is long compared with the fluorescence decay

time, and the emitted light will be plane-polarized, parallel

124 to the polarized excitation light. A small molecular fluorophor

causes depolarization because its rotation relaxation is faster

than the fluorescence decay time, and thus, its rotation may

change between excitation and emission; light will therefore be

emitte mann i d y different directions (i.e., light wile b l

depolarized). However, if such a small molecule is attached to a macromolecule a fluorophor-labele n i s a , d antigen-antibody

complex, the small fluorophor will behave similarly to a large fluorophor and emit polarized light. This phenomenon can be used analytically r examplefo , n itnmunoassai , y e methodth r fo s quantitation of a number of therapeutic drugs, drugs of abuse,

and other small molecules. The central principle of these method s competitioi s n betwee e analyte sampla th nth d n an i e known amount of fluorophor-labeled analyte for a limited number

of binding sites on an antibody to the analyte. The amount of analyte present in the sample will, therefore, have a direct effect on the change in fluorescence polarization. For

example, if the amount of analyte in the sample is small, more

fluorophor-labeled analyte- (antigen-) antibody complex forms e degreanth d f depolarizatioo e s smalli n . Converselye th f i ,

analyte concentration in the sample is high, less fluorophor-

labeled analyte- (antigen-) antibody complex formsd an ,

depolarization is significant.

Immunoassays that use labels such as enzymes and

fluorophores described abov e know ar es nonisotopi a n c immuno-

assays. These assays have comparable sensitivity with radio-

immunoassay and have the additional advantages of having stable

reagents and simple waste disposal of spent reagents. Also,

these assay e amenablar s o automationt e d instrumentatioan , n

125 exist r thifo ss purpose r exampleFo . , centrifugal analyzer] [6 s

e commerciallar y available that incorporate both enzymd an e

fluorescence immunoassays as part of their repertoire of auto-

mated chemistries. These machine e eas o operatear st y , need

only microliter volume f boto s h reagent d samplesan s , allor fo w

calibrator e processeb o t s n reai d l time with samplesd an ,

provide for on-line data acquisition and processing. In addi-

tion, special purpose instruments are available for automating

fluorescence and fluorescence polarization immunoassays [5].

In summary, immunoassays are a very popular and widely

used technique to measure a large variety of analytes in

biological fluids. They offer excellent sensitivity and speci-

ficit d hencan y e their wide applicability A variet. f labelo y s

e usear d with immunoassays; however o disposat e du , l problems

with radioisotopes, radioisotopic-based assay e rapidlar s y

being replaced with nonisotopic labels such as enzymes and

fluorophores.

REFERENCES

[1] BERSON, S.A., YALOW, R.S., Quantitative aspects of the

reaction between insulin and insulin-binding antibody,

J. Clin. Invest. _3_8 (1959) 1996.

] Clinica[2 l Immunoassay Instrument MarketU.S.e th n ,i s Fros t

and Sullivan, New York (1984).

] KOHLER[3 , MILSTEING. , , "ContinuouC. , s culture f fuseo s d

cells secreting antibod f predefineo y d specificity,"

Nature 256 (1975) 495.

[4] RUBENSTKIN, K.E., SCHNEIDER, R.S., ULLMAN, E.F.,

Homogeneous enzyme-immunoassay w immunocheraicane A . l

technique. Biochem. Biophys. Res. Commun 7 (19724 . ) 846.

126 ] JOLLEY[5 , M.E., STROUPE, S.D. al.t ,e , "Fluorescence

polarization immunoassay I." Clin. Chem. 27_ (1981) 1190;

II. Clin. Chem. 27_ (1981) 1575.

] BURTIS[6 , C.A., TIFFANY, T.O., MROCHEK, J.E., "The

centrifuga n clinicai e lus analyzels it analyses, d an r "

Centrifugae Methodth r fo s l Analyzer (SAVORY, ,J.

CROSS, R.E. Eds), AACC Press, Washington, D.C. (1978. )3

DISCUSSION

u thin. BraunT yo woult i o k D e possibl.b d o fint e a heptand e formina g complex with some inorganic trace component and using the complex as a basis n immunoassaa f e o determinatioth r fo y f thao n t inorganic trace component? t personall no e possible b o d . y BurtisI C ma t y effortt bu I ,an kno . f o wo t s develop suc n assaya h I doub. t that suc n assaa h y s sensitivwoula e b d s a e spectroscopic methods n additionI . n immunoassaa , y woula singl e b d e element determination while spectroscopic methodmultielemente b n ca s woulI . d also think thae spectroscopith t c methods woul e methodologicallb d y simpler than a n immunoassay.

H. Ross. Can immunoassay techniques be effectively extended to the analysis f materialo t associateno s d with clinical interest, i.e., organic pollutants t environmentaa l levels?

. BurtisC . Yes n antiboda e prepared.b f I n ca y n immunoassaa , e b n ca y developed r exampleFo . , many such assays have already been developer fo d polynuclear aromatics and their metabolites. Antobodies for smaller molecular weight compound e prepareb n ca sy bindin b d e compounth g a large o t d r molecular weight compound and then injecting the combined molecule into an animal which then produce n antibodya s .

R. Rosenberg. DoeA havy futureRI san e ?

. BurtisC y opinionm n I .A wil RI , l have only smal e lclinica th rol n i e l laboratory of the future. Since non-isotope labels (primarily enzyme and

127 fluorescence labels) are available and methods based on them can be easily automate n existino d g instrumentation. These type f assayo s s will continuo t e displac d replacan e A techniquesRI e , however r researcfo , h purposes I thin. k RIA will continue to be used. As I mentioned in my talk, the major disadvantage with RIA today is the growing problem of disposing of the low-level wastes thae generatear t d wituses it h .

128 ROLE INATH F E O COMPARE S AA O DT CONVENTIONAL METHODS OF ANALYSIS FOR GEOLOGICAL SAMPLES IN CANADA

E.L. HUFFMAN McMaster University, Hamilton, Ontario, Canada

Abstract

The instrumental neutron activation analytical technique (INAA) both competes witd complimentan h s conventional commercial analytical methods like X-ray fluorescence emission spectrometry (XRF), direct coupled plasma emission spectrometry (DCP), inductively coupled plasma emission spectrometry (ICP), inductively coupled Dlasma-mass spectrometry (ICP-MS d atomian ) c absorption (A.A.). The major advantages to the use of INAA in the geological field includes the fact that the analysis can be done non-destruc- tively witnout encountering tne many problems innerent in acid dissolution or fire assay techniques. The sensitivity for many elements are still unrivalled by commercially available analytical techniques for gold, platinum-group-elements, many of the rare earth element otned an s r elements multi-elemene Th . t naturf o e the analysis comoined with eas f automationo e , data collectiod an n soectral analysis procedures allows for rapid turnaround of samplecostw lo t . a s This technique competes very favourably with virtually all other analytical techniques. The geological market being serve y INAb d A includes univer- sitd governmenan y t researcher e minera th s wel a s sa l l exploration industry neede f Th theso s. e groups varies consideraol wild an yl De discussed, "The various applications of INAA to the geological community will als e discussedb o .

MINERAL EXPLORATION

The requirements of this sector for an analytical method includes providing high quality data with a fast turnaround time at low cost. The field exploration seaso n Canadi n d aan y beginMa n i s usually extends into October with some overburden and diamond drilling also occuring from December to e shor th a resulApril tf s o A fielt . d seasoe th n explorationist usually wishes and expects to receive analytical results within a few days and up to three week f o submissios f o samplesn e choicTh f o . e analytical metho s i dals o frequently determiney b d the elements and sensitivities available for each different analytical technique, as well as cost. In the last few years, mineral exploration b v tne major

129 mining companies has been depressed as a result of low metal prices. The abundance of laboratories serving this sector as well as the more ubiquitous f o automate e us d multielement analytical techniques has caused severe price competition with price per element now probably at 50% of prices five years ago. There are two qualities of analyses being sold by Canadian laboratories. These are classed as assay (accuracy and precision of better than + 2-5%) and geochemical (accurac d precisioan y , 30%)+ f o ne .Th e sampleth bul f o sk being analyze r thifo sd sector are of the geochemical variety. The lowest cost activation analysis packages typically will give better precision and accuracy than the required + 30%.

The choice of analytical technique is determined to a large extent by the elements being sought. Exploration in Canada has for the last few years centered almost exclusivel n o goly d with more minor exploration for platinum group elements (PGE), base metals and exotic elements. Prior to 1981 the bulk of the exploration was related to uranium.

In general there are certain elements or group f o elements e requireminerath y b ld exploration sector which are determined accurately and at low cost by techniques other than INAA. These include the major rock forming oxides (Table I) and some trace elements whice determinear h d primarily b y X-ray flourescense (XRF), inductively coupled plasma r o (ICPdirec) t coupled plasma (DCP) emission spectrometry techniques. For this type of analysis, INAA can't compete either in cost or precision for many of the elements. Another group of elements which are better determined bv other than INAA

Tabl I eMajo d soman r e trace element analysi y X-rab s y fluorescence

Major Oxides Trace Elements (Detection limit (Detection limit; 0.011) of 10 ppm)

Si02 Fe203 Rb

A1203 MnO Sr

CaO Cr203 Y MgO P2°5 Zr Na20 Ti02 Ba

K20 Loss on ignition Nb

130 techniques are what is termed a base metal suite which is shown in Table II. This suite includes some g whicA e verd ar han y d C , elementNi , Cu s lik, Pb e difficul o t determint y b INAn ei geologicaA l material e leveth t la s require r geochemistryfo d .

Table II Base metals are usually determined by ICP, DCP or atomic absorption techniques A typical DCP package is shown below.

Detection Detection El ernen t Limit Elemen t Limit

Cd 1 ppra Mn 2 ppm Co 1 ppm Mo 1 ppm Cu m pp 5 . 0 Ni 1 ppm Fe 2 ppm Ag 0.5 ppm Pb 2 ppm Zn 0.5 ppm

This paper wile devoteb l o thost d e elements which are done exceptionally well by INAA. Some of these element e goldar s , platinum group elements, rare earth elements s wela ,s othersa l . Geochemical Method r Golfo s d e Platinuanth d m Group Elements

Various methods exist for the analysis of gold e b broke n ca n d dowan n e followinintth o g main categories :

a) Instrumental determination by a very sensitive analytical technique like instrumental neutron activation (INAA). b) Fire Assay Collection (usually lead collection with silver added as a collector) followed by: i) gravimetric finish -(dissolving silver from the precious metal cupellation bead and weighing the minute gold (+PGE) flake). This is the classical technique. 11) instrumental finish by INAA, or dissolution of the cupellation bead followed by atomic absorption, flameless atomic absorption or DC plasma emmission s pec t rorne t ric techniques. ) c Acid dissolutio e samplth f o ne followey b d solvent extractio n instrumentaa f golo d n an d l analysis of the organic solvent by atomic absorption C oplasmaD r A subse. f thio t s technique, much less commonly use s direci d t cyanide extraction whicn ca h replace acid dissolution. Solvent extraction is an essential step. Direct analysis of an acid digest raav give rise to s pec t romet ric interferences and therefore erroneous gold values.

131 OJ to Table III Methods of analysis for gold at NAS •OÛH13 H SAJ1I 1 f CRU AI.'AI Kill M AflAH l Y l »OUI IN' f UMIIir HE IEHIIINE.i [lulL E l1 l I » niMHf N1S HAU RIAl REQUIRI U ii m uni u i Ul 01 r. IMUI IANEMISI Y1 m (AUm i ) SAJ1PU

HuU Millin& y in h g us i L 1 1 ANA 20y No 1 pplj detection Huit a (unction of fire «ssay reagent blank g Crushinlu 1 1 l M 1 g 2 PKEIRRAII fANA '9 No 0 1 ppb used to determine background levels In rocks y m 1 1 i H 1 y in i l s Cu r 3 Direct Inad INAA 30g Yes, r.roii1 p S ppb sample sizes can vary from g y Cm r 1 usluii1 i M 1 y 4 Ni Su 1 phi Oe 50g Yes, Rh.Pt.Pd.Os.Ru. 1 ppb use n evaluatinI d depositE PG g s FA INAA li .Ri- Crushing 1 Mi 1 1 ing 5 fA DCI' 20g res, Pd.Pl 2 ppb use r routin(o d exploratio E PG e n Soil Drying & Screening I r ANA 20g No 1 ppb amount of organlcs luvt to Method * Interpretation Dry ing & Screen ing 3 Direcd a r I I t INAA 30g Yes. Grou1 p 5 ppb Dr y l ng & SL r eeit my 5 r /\ »tr 2og Yes, I'd.l'l ? I'PI. e>nlor*tloE usefuPG r fo l n r.iacidn u l Heavy Mineral 6 Oued Irrad 1g to Ye s, Grou r o Grou 1 p 2 p 5 ppb f I samples tus « reuseb t n timeli d y fashion Grou? p 01 Penned Separation INAA 609 elements only are possible The whole concentrtte Concentrate (usually non-nagne ) shoul c i t e analyseb d d Sepei UK ? Une; t Inad lg to Yes , Group 1 2 ppb lower detectio b possiblpp n 1 limit y 0 b e o t s -250 mesh INA* 30 ) using metho2 d Humus (Hull) Drying. Macérai mg. 8 Direct Irrad «9 Yes, Grou3 p 1 ppb sample mis e 90*b t * .organi c Detection Unit Briqug m 1 e1 INAA affecte y inorganib d cs végétât ion Drying. Macérâ. mg t 9 Direct Irrad 8g Yes, Group 3 b pp 1 0 detectio ne slightl b limi y M t y elevateo t e du d Briquet t mg INAA high levels of Na or Br in some species Drying, Ash mg 10 Direct Irrad «Ig to Yes, Group 1 5 ppb ashing losse e possibleir s , ashin s expensivI g t bu e INAA 29 advantage of using larger sample Ash /leid I-5I Water Concentraten o d 11 Direct Irrad 100 ml Ye S , Group 1 1 5 ppt caution mus e takeb t n collectini n g s samplI u (A e Activated Charcoal INAA adsorbed onto polyethylene) Ivaporat ion 12 Direct Irrad 100 oil Ye , s Grou1 p 1-5 ppt evaporation In polyethylene 1 triers INAA

lakg n e1 1 botto 1 l M m& g In y Or d a i 1r I 3 1 c Ore i 5 - 30.J Ye s, Grou1 p 1-5 ppb detection limit a function of the Inorganic content Sediments INAA

Drying ( Mill Ing, H Direct Irrad «g 1-5 ppb can only be done If the organic content Is high Brlquetting INAA g m 1 Dryin l Mi k g 15 AN A r S ?0g 1 ppb If organic content is very high the sample may have e asheb o t d (ashing losse y resultma s )

(,ruup 1 llier-rul ar.11 vil ion elements Ag.Ai ,Ba ,Ca ,(o ,Cr ,Ee ,llf ,Mu.Na ,Ni ,Sb,Sc .Se ,la ,1h,U,W,Zn ,la .Ce .Sm.tuYb.Lu Oroup 2 [pitliermal activation elenrents Ai ,8a .Co.Cr ,Ee ,lto ,Sb,Sc ,la ,Th.D,H,la Group 3 lluiiiui or VeyeUtiun Ag .As ,Ba ,Br ,(.1 .Co ,le ,IV),Se ,Sb ,1 a ,Ili ,U,W,Zn All of these techniques have advantages or disadvantages depending on the type of material being analyzee requireth d an d d geological information.

Analysis of Gold by INAA

A wide variety of techniques employing INAA are available depending on the type of material to be analyzed, and the type of information required. These techniques are summarized by method number (Table III) which will be referred to again. A description of the relative merits of each of the technique s i describeds , accordin e typth f o eo t g sample media.

Rock and Soil

Various neutron activation analysis techniques, which yield gold detection limits from 0.1 ppb to 5 e ppblistear , s Methoda d s 1-5. These include both fire assay (FANA) and direct irradiation based techniques. Sample aliquots range from less tha1 n 0 grams50 e mos Th grao t .tm popular procedure (Metho ) involve#1 d a slea d fire assay collection o n 20 gram aliquots followed by INAA of the fire assay bead. This techniqu a detectio s ha e b npp limi1 f o t which is a function of the fire assay reagent blank. The main competing techniques would involve the same fire assay procedure but followed by acid dissolution n ana analyticad 2 ppbr o L ^ (D l P finisDC y b h flameless AA (DL l ppb).

In many instances a researcher may want to look e differenceth at r examplefo s , betwee n alterea n d versu a sfres h volcanic rocr backgrounfo k d levelf o s gold. Metho 2 woul# d e applicablb d r thifo e s typf o e survey e disadvantagTh . o thit e s technique th s i e relatively small sample whic s analyzei h a resul s a d t e radiatiooth f n hazard from larger samples. Gold contamination is not a problem as the sample is irradiated prior to fire assaying, and regardless of how much gol s i dintroduce d after irradiation only the radioactive gold e wilb measuredl . This technique is described by Bornhorst et al, [1]. Metho 3 describe# d e INAth s A technique th r fo e analysi6 elemen2 a f o ts package including goldA . 0 gra3 m sampl s routineli e y encapsulated, irradiated and afte a suitablr e delay, measured. o n Ther s i e chemical handling of the sample. This package allows one to obtain statistically meaningful gold values on relatively large sample sizes at the low ppb level as well as many other elements simultaneously, nondest ructively, and at little additional cost to the gold analysis itself. A wealth of geochemical and geological information is obtained. Included in

133 this package are:

a) Au to 5 ppb This detection limit is suitable for most exploration surveys. m Thipp 2 s grou f o elemento t p e b s A n ca ) sb Sb to 0.2 ppm useful as pathfinder elements W to 4 ppm which may occur in many types of Ba to 100 ppm gold deposits. Negative drill Mo to k ppm results may occur due to the m inhomogenoupp 5 o At g s distributiof o n m goldpp 0 5 . However o t n Z a mineralize, d e followe b m zony pp ma 5 e d usin o t ge S these element e mineralogth e f i sth f o y deposi s suitablei t .

1 Thes1 e o elementt a C s) c will reflece th t Na to 0.05 % degree of car bon 11izat ion or serici tizat ion , respectively, when use n conjunctioi d n with other elements listed in this package. m Thespp 1 e rar o t e a eartL ) hd elements (REE) e considerear m pp 3 d relativel o t e C y immobile m durinpp 1 g0. mos o t t m alteratioS d an n Eu to 0.2 ppm me t amorphic processes. Ratios m suc pp s La/Sm a h2 0. , o YLa/Lt b d Sm/Euan u , Lu to 0.1 ppm combined witn the absolute abundances, orovide valuable information on rock tvpe, degree f differentiatioo e parenn t f o nt magm d potentiaan a r oasfo l e metal deposits \ descriptio. e th f o n e rarth ef o earth e d otneus an s r elements in t nis 26 element package are given ov Campbell et al, [2].

e) Cr to 10 ppm These elements will help identify Co to 5 pptn rock type (le, ultramafic bodies Ni to 200 ppm will be higher in Cr, Co and Ni) Fe to 0.02 % and might help locate potential platinum group element deposit host rocks. m Sompp f thes1 o e e o elementt a T e ) ar sf U to 0.5 ppm considered relatively immobile m (evepp 5 n 0. mor o o thas e REEet th d nh T an ) Sc to 0.1 ppm will help identify rock types. Hf to 1 ppm Commercial quantities of Ta and U may als e locatedb o . Correlation of rock units from drill hole to drill hole will also be aided by usin e gpreviou th thes d an e s group of elements.

134 Method 4 is a nickel sulphide fire assay procedure followe y b INAd A analysi e residuth f o s e forme n o filterind e concentrateth g 1 solutioHC d n which was formed on the dissolution of the nickel sulphide button. This method has been described by Huffman et al, [3]. Although this method provides excellent data on the whole platinum group this method is relatively expensive as it requires two irradiations and three separate counts. A more cost effective method of geochemical exploration utilizes the lead fire assay collection followed by a DCP s i liste d s an Methoa du A . 5 d d an d P , finisPt r fo h In the event PGE mineralization is found it would then be advantageous to analyze for the whole platinum group.

Glacial til r heavo l v mineral concentrates

Overburden drillin a preferre s i g d techniqur fo e gold exploratio n glaciatei n d areas. Heavy minerals e concentratear de non-magnetith froe tild th man l c fractio s i usualln y analysed. Commonly, heavy mineral concentrates weigh from 1 to 25 grams with the occasional sample weighin 0 s grams6 a mucg s a h . The gold "nugget" effect requires that the whole sampl e b analyzee r feaw f fo o missindfe r e th g particles in the sample that may be in the unanalyzed portion. In a 10 gram heaw mineral concentrate, a singl 0 micro20 e n flak f golo e d would generat a 500e 0 ppb gold result which is considered highly anomalous. If e thiflak on s sincludewa ea sampl n i ed split e b t analyzet no no e whicsampls th y wa dhma e recognize s anomalousa d . Fire assayin e samplth f o eg requires additional grinding, further enhancine th g possiblit f losseso y . Whee entirth n e sampl s usei e d for fire assaying o materiathern s i e l availablr fo e raineralogical examination or additional analytical testing.

The INAA technique (Method #6) is the answer to all these problems. The heavy mineral concentrate is encapsulate o grindin(n d g required), irradiated an d measure6 element 2 s man a s a r y fo ds including gold. However, with this technique e probleth , f selo m f shieldin f golo g d arises. This effec s i tcause y b d the high nuclear cross section of gold. There is a depression of the gold value reported as a result of large gold particles e absorbinneutronth n l o al s g the outer layers of the particle. The effect has been calculated to cause an apparent decrease in gold content of 5% for a 200 micron diameter sphere and 9% for a 400 micron sphere. Any flattening of the particle which generally occurs in nature should reduce this factor. Reported values will stile b l considered highly anomalous due to the large grain sizes involved and this depression should not concern most exploratlonist s.

135 Some exploration programs have used the clay size e tilfractioth l f o (-25n 0 mesh r golfo ) d exploration e applicablar .7 r o Method o 2 t e , 1 s this portion of the sample and the selection would e requiredepenth n o d d detection limits.

Vegetatio d humuan n s sampling

In area f o sdee p overburde e geologisth n r o t geochemist has frequently turned to analyzing soils in the search for gold deposits. In areas of transported overburden thers i frequentle o n y relationship between mineralized bedrock and overlying soil. Even in areas where soil may reflect buried mineralization, anomalie e verb y y ma serrati c due to the "nugget" effect of gold. Arid to semi-arid climates frequently have very sparse vegetation cover, howeve e specier th man f o ys present may be phreatophytes (ie, roots may reach depths in excess of 100 metres) as a result will sample much greater depths than a soil sample.

Unless a suitable precipitation mechanism exists, gold complexes will teno t remaid n i n solution in the groundwater yielding no geochemical anomal n soili y r tilo s l adjacen o mineralizationt t . o t theiDue r extensive root systems, trees effectively sample large volume f o d ssoian l groundwater. Many species may be cyanogenic or produce a very acidic root environment which has the ability to dissolve gold particles and then absorb gold into the vegetation.

Before the late 1970's, there was not a suitable economi d productivan c e analytical technique which could detec b rangetpp golw . lo d e contentth n i s Now, detection limits of 0.1 ppb gold can be achieved using direct irradiation INAA (Method #9). The vegetation is dried, macerated and 8 grams is compressed, and analyzed in a multielement mode bv INAA. Further detail d limitationan s e availablar s e from Cohen et al, [4] who conducted a b iogeochemist r y survey over the TeckIn t erna tional Corona and Noranda Golden Sceptre deposits in the Hemlo region of northwestern Ontario, Canada. This newlv discovered inning camp will become one of the largest gold mining areas in North America.

Other researchers believe that ashin e samplth g e and analyzing the ash (Method #10) will allow analysis on a much larger sample. Caution should be exercized as there may be extreme ashing losses and e chanctn f o contaminatioe n increases with added handling. There will usually be a substantial asning charge from commercial laboratories as well as a relatively slow turnaround time. The ash will generally be analyzed by INAA.

136 The analysis of humus continues to be a favourite metho r golfo dd exploratios i d an n described under metho . 8 d This techniqu s i more e fully describe y Hoffmab d d Brookeran n , [5].

Wa t er

The analysis of water for gold by INAA has been described by Hamilton et al, [6] and shows great promise. This technique developed in Australia has not received much attention yet from exploration geologists t leasa ,n Canadai t . With this technique s iveri t y important tha a tpolyethelen e liner and/or activated charcoa e useb t a d ) l12 r (Methodo 1 1 # s the time of sample collection. This is important as the gold will absorb onto the sides of polvethelene or glass bottles.

Lake Bottom Sediments

This technique has been used in both regional and property scale surveys e detaileTh . d property scale survey has been very useful in helping to evaluate geophysical conductors which are found under water. The survey is conducted by boring holes throug d dredginan h e lak ic p laku eg e bottom samples usin a G.S.Cg . tvpe lake bottom sampler.

The samples obtained are difficult to analyze becaus ef o organice the usuallar x d ymi an a s y inorganics. INAA Methods # 13 to 15 can be ased dependin e specifith n o g c e samplenaturth f o e .

UNIVERSITY RESEARCH

e worTh k performe r thifo ds sector usuall/ require e highesth s t precisio d accuracyan n . Cost and turnaround time are of lesser importance. Generally speaking university researcher e trvinar s g to obtai s muca n h informatio s possibla n a sampl n o e e and usually require the best analytical method for each element.

NAS provides various multielement packages with 2 element5 o ut pd optionan s s utilizin e followinth g g analytical techniques: INAA, Delayed Neutron Counting (DNC), Prompt Gamma Analysis, A.A., DCPF XR , d an Inductively Coupled Plasma-Mass Spectrometrv (ICP-MS). Three different grade f analyseo s e ar s available, however universities prefe o purchast r e the research grade shown in Table IV. This qualitv of analysis will allov, plotting of chondrite normalized diagrams essentia r fo helpinl o t g interpret genetic history of rocks.

Various options which are frequently requested include boro d gadoliniuan n m determination v promob s ^ gamma analysis, and uranium analysis by DNC.

137 Table IV Multielement multi-method package by INAA, DCP, AA and XRF (Total Metal)'.

DETECTION LIMITS DETECTION LIMITS Element Basic Exploration Research Elément Basic Exploration Research

*9 ppm 0.5 05 05 Mg% 001 0 01 AI °* - 0 01 001 Mn ppm 20 2 2 As ppm 5 2 1 Mo ppm 10 5 2 Au DDD 50 20 5 ' a N H 005 001 001 8 ppm - 10 10 Nt> ppm - 20 10 8a ppm 1000 20 10 Ni ppm 5 1 1 Be ppm - 10 1 P «,» - 001 001 e< ppm - 0 5 1 0 Pb ppm 5 2 2 Br ppm 5 1 05 Rt> ppm 100 20 10 Ca % 5 0 0 01 0 01 Sb ppm 1 02 1 0 Ca ppm 2 02 02 Se ppm 10 3 os Co ppm 5 1 1 0 Si

Rar» Earth and Actlniide Elementi

Se ppm 0 5 0 1 001 Tb apm - - 1 0 La DP m 05 0 1 Y6 ppm 05 2 0 0 05 Ce ppm 5 3 1 Lu ppm 2 0 0 05 001 NO ppm 10 5 3 U ppm 5 05 1 0 Sm ppm 0 5 0 1 001 Th ppm i 05 02 Eu opm 5 0 2 0 0 05 Dy- ppm 0 5 5 0 05

The entire platinum group element family is determined by nickel sulphide fire assay followed by INAA analysis of a residue produced during the processing of the sample. The detection limits achievable are shown in Table V.

Table V The platinum group elements, rhenium and gole determinear d a combinatio y b d f o n nickel sulphide fire assay collectio- n acid dissolution and INAA analysis of the residue. The following detection limits are achievable on most materials.

Detection Detection emenl E t Limit Element Limit

Rh 1 ppb Os 3 ppb Pd 5 ppb Ru 5 ppb Pt 5 ppb Re 5 ppb Ir 0.1 ppb Au 1 ppb

138 Competing techniques for the rare earth elements includes mass spec tromet d ICP-MSan , ry XRF P IC ., Isotope dilution mass spectrometry provides an excellent qualit f dato y a howeve e methoth r s i slod w and expensive. Spark source mass spectrometrv is also sensitive howeve s i alsr o expensiveF XR . analysis suffers mainly from limited sensitivity. P analyseIC e rarth er fo searth s suffers e froth m disadvantage of putting the sample into solution. This technique also frequently lacks sensitivity for some of the important rare earths like Eu. ICP-MS is new technology which links an ICP front end to a mass spectrometer. Preliminary experimentation shows much better sensitivities for most rare earths than the P techniquIC e howeve e samplth r e mustt pu stil e b l into solution. Sensitivit r certaifo y n rare earths e stilar lik u lE e bette y ÏNAAb r .

The whole PGE family is commonly analyzed bv a variety of techniques following the nickel sulphide collection. Most of these techniques for example, , suffeRu distillatio d r an fros O m r fo beinn g very costly, slod witan w h poor sensitivity e ICP-MTh . S technique shows promise particularly for Pt and Pd bus stili t l largely experimental.

The main competitive technique for uranium analysis is flourimetry. Flourimetry suffers from bein a mucg h higher cost analytical technique than delayed neutron countin s a wels ga havinle th g requirement of putting the sample in solution.

Boron and gadolinium by prompt gamma analysis has the advantage of not having to put the sample into solution as well as having increased sensitivity compared to other techniques like DCP. This is important because the boron is usually tied up in very resistate minerals like tourmalinest whicno o d h readily dissolve.

GOVERNMENT GEOLOGICAL SURVEY APPLICATIONS

This market sectors' needs lie between that of the universities and the mineral exploration community. Generally, research related to mineral deposit r relateo s o formulatiot d f genetio n c models will require the best sensitivities similar to the needs of universities. The government geological survey n i sCanad a both federa d provinciaan l l also conduct regional geochemical reconnaisance programs. The aim of the latter type of program is to develop databases which will be of use to the mineral exploration community. This type of survey typicallv will consis f mano t y hundred r thousando s f sampleo s s e analyzeanar d r selectefo d d specific elements. Usuall e elementth y s requested wil e determineb l V D d e analyticath l techniqun i e e us whics beeha n i h n

139 previous similar surveys regardless of whether it is the best technique. For most of these surveys uranium is usually determined by delayed neutron countin d heavan g y mineral e determinear s y INAAb d . One of the provincial surveys is also analyzing lake bottom sediments in a multielement mode by INAA, as they believe they can obtain the most information at the lowest cost with this technique.

Conclusions INAA compliments other analytical techniquen i s that the elements it can provide cost effectively are frequently those elements which are difficult to analyze by competitive analytical techniques. The major multielement analytical techniques used most ofte y b laboratorien s providing e serviceth o t s geological community include INAA, ICP and DCP emission spectrometr d an X-ray y fluorescence spectrometry. Certain groups of elements like the major rock forming oxides and base metal suites described are better determine F y plasmtechniqueb dXR r o a s thay b n INAA. This primarily because of sensitivities and cost of analysis. Analysis by INAA, DNC or prompt gamm e preferrear a r goldfo e platinudth , m group elements, selected rare earth elements, uranium, boron and several elements which are usually included in multielement suites like Ta, Hf, Sc, Cr, Br, I, Cs, W, Re and In. Ia commercian l sense s a INAs provee ha b A o t n cost effectiv s a othee r analytical techniqued an s provides usually a similar turnaround to potential competing techniques.

References

[1] BORNHORST, T.J., ROSE, W.I. JR., WOLFE, S.P. and HUFFMAN, E.L., Gold conten f eleveo t n french geochemical reference samples. Geostandards Newletter, _3 1 (1984) 1. [2] CAMPBELL, I.H., LESHER, C.M., COAD, P., FRANKLIN, J.M., GORTON, M.P. and THURSTON, P.C., Rare earth element mobility in alteration pipes below massive Cu-Zn sulphide deposits. Chemical Geology, 45 (1984) 181.

[3] HUFFMAN, E.L., HANCOCK, R.G.V., NALDRETT, A.J., VAN LOON, J.C., and MANSON , A., The determination of all the platinum group elements and gold in rocks and ores by neutron activation after preconcentration by a nickel sulphide fire-assay techniqu n largo e e samle sizes. Anal. Chem. Acta 2 (197810 , ) 455.

140 [4] COHEN, D.R., HUFFMAN, E.L. and NICHOL, I., Biogeochemist r y : A geochemical method for gold exploratio e Canadiath n i n n shield. . J Geochem Exploratio n press)(i n .

] [5 HUFFMAN, E.L d BROOKERan . , E.J., Biogeochemical prospectin r golfo g d with referenc o somt e e Canadian gold deposits. Rubey Special Volume 5, Chapte , Minera9 r l Exploration: Biological Systems and Organic Matter: Carlisle, Berry, Waterso d Kaplaan n n eds; Prentice Hall (1985).

] [6 HAMILTON, T.W., ELLIS , J. FLORENCE, , T.Md .an FARDY, J.J., Analysi f golo s n surfaci d e waters from Australian goldfields n investigatioA : n into direct hydrogeochemical prospecting for gold. Econ. Geol. . (198378 , ) 1335.

DISCUSSION

R. Rosenberg . e typicath Whic e ar hl irradiatio d measuremenan n t times used for NAA.

E.L. Hoffman. The typical irradiation time for long-lived isotopes would generally not exceed 1 hour for a 0.5 gram sample for geological rock material 12 -2 -1 usin a gtherma l neutro 0 1 n.cx n . 7 flum Countin f S o x g times would vary from 200 to 2000 seconds. Occasionally we may irradiate samples up to 10 hours for a specific reason.

141 DETERMINATION OF MINOR AND TRACE ELEMENTS IN DIETS USING NEUTRON ACTIVATION, INDUCTIVELY COUPLED PLASMA ATOMIC EMISSION AND ATOMIC ABSORPTION SPECTROMETRY

R. SCHELENZ Departmen Researcf to Isotopesd han , International Atomic Energy Agency, Vienna

Abstract

The recognition of the role played by trace elements in human and animal healt s stimulateha h d interes n establishini t g their concentration a wid n i es variety of biological materials. One of the current topics of interest in human environmental and nutrition research has to do with establishing recommended dietary allowance r essentiafo s s a wels la provisionall l y tolerable weekly intake r toxifo s c elements e migratioTh . f traco n e elements in air, wate d soi an rs reflectei l e foodchaith n i d f mano n e minera.Th d an l trace element content f individuao s l foodstuff d especiallan s y human total diets are promising tools to detect possible health hazards in the environment.

In this connection there is a need to determine the concentrations of elements in foodstuffs, especially in total diet specimens. An important requirement in such work is the application of appropriate analytical quality control procedures f baseo certifie e d us partiall e d th referenc n o y e materials. Although several biological material f environmentao s n i l w interesno e ar t common use no human diet is available that can be regarded for checking the accuracy and precision of analytical methods for the determination of trace and minor elements in these specimens. o receivT e nutritional data with respec minoo t d trac an r e element contenf to total diets an Agency Co-ordinated Research Programme (CRP) was created joinin 3 laboratorie1 g e Membeth f ro s State l ovee worldal e sth r Th . Chemistry Uni f Agency'o t s Laboratorie t Seibersdora s f serve programme th s s a e a Reference Laboratory.

In this context, a co-operation with the National Bureau of Standards (NBS) e Departmenanth d f Agriculturo t e Uniteth f do e States (USDA), Human Research Centre Beltsville was established to produce freeze dried and powdered total diet materials whice assumear h o servt ds a secondare y standardr fo s calibration purposes and to improve the accuracy of analytical trace element

143 result n suci s h matrices e AgencTh . y contribute a finisd h diet S (H-9)MB e ,th a mixed American diet consistin f abou o g0 individua20 t l food items (US-Diet 1) and the USDA a typical American diet (TDD-1D).

The results of 21 (AI, Au. B. Br, Ca, Cd, Cl, Co, Cr, Cu, Fe, Hg. K. Mg, Mn, ) elementZn , Se s, determineRb , P , Mon thesNa i ,d e diets using instrumental and radiochemical neutron activation analysis (INAA, RNAA), inductively coupled plasma atomic emission spectrometry (ICP-AES) and atomic absorption spectrometry (AAS) wil e give b ld compare an n d with respec o accuract t d an y precision as well as with the results obtained by the 2 other collaborating laboratories. Experiences in using Hydrated Manganese Dioxide (HMD) for the radiochemica n totai e lS died an lt r sampledeterminatioC , Cd s , afteAs f ro n separation from Na, K and P will be reported. The advantages and disadvantages of these analytical methods will be discussed.

DISCUSSION

W.D. Shults e thre. th Whaee referencar t e diets whic u mentionedyo h e Ar ? they comparable?

R. Schelenz. These diets have been produce n e IAEframa th f An o ei d Coordinated Research Programme (CRP) on "Human Daily Dietary Intakes of Nutritionally Important Trace Elements". These diets are comparable to such an extent that they represent typical diets related to their origin which not necessarily means that they are also comparable with respect to their minor and trace element content.

. RosenbergR e contenS Finnise th e f Th o t.h diet shoul e loweb d r than thaf o t the U.S. diet. Isn't that true?

R. Schelenz. As a matter of fact the Se content of the two U.S. diets (TDD-1D, U.S. Diet 1) is two times higher than for the Finnish diet (H-G) as indicated in the tables. [The Se concentration of plant material reflects the respective soil Se concentration). This is probably due to the fact that the average Se content of U.S. soils is significantly higher than found in soils originating from Finland becaus f concentrationo e .

144 V. Krivan. I am surprised about the almost continuously higher values for iro y INAAb n . Possible primary interference nuclear reaction n easilca s e b y located, and so a systematic error analysis must relieve the source of this error I shoul. d therf lik i e som k ear e as eals o approachet o e th n o s development which would allow to specify the valency state of elements in such a complex material as the different diets are representing.

R. Schelenz ) Possibl(1 . e interference f nucleao s r reactions have been considere d don'an d t give suppor o significant t t highey m r o valuesT ) (2 . knowledge only a few attempts have been made to approach this problem because a total diet composite sample represent a vers y complex matrix.e th Mos f o t analytical method r tracfo s e element determination n biologicai s l material used a wet digestion procedure. Therefore the probability of measuring speciation artefact s increasingi s .

C.J. Pickford. Are total Co and Fe contents sufficient: surely organically bound values are more important toxicologically.

145 PRESEN FUTURD TAN E PROSPECT NEUTROR SFO N ACTIVATION ANALYSIS COMPARE OTHEO DT R METHODS AVAILABLE

. REVEG L Laboratoire d'analys r activatioepa n Pierre Sue, Centre d'études nucléaires de Saclay, Gif-sur-Yvette, France

Abstract

e evolutioTh analysif no Neutroy sb n Activation during the last fifty years is briefly described. The reasons for its persistant success, in spite of more recent methods are examined. These reason e specifi e baseth sar n o d c advantagef so neutron irradiation : neutron penetration into the material, sensitive detection of 60 elements, selectivity and reliability. Furthermore the sensitivity is inversely proportional to the natural abundanc elemene th f eo t (fo 80)> rA . Howeveo t r benefit fully from the advantages of N.A.A., it is essential to study the potential sources of error inherent in each step of the analysis (sampling, preparation, calibration, irradiation, chemical separation, radioactive measurement, calculatiod an n interpretatio results)e th f o n . These riske sar discusse d examplean d s given e principaTh . l arean i s which the N.A.A. is being used more and more are ultra-traces analysis, studies on the multivarient correlation between elementvalidatioe th d san f o n other methods. Based on present applications and the likely progres f thiso s technique s futurit , e prospect e discussedar s .

I - INTRODUCTION

e firsTh t results obtaine neutroy b d n activation analysis (NAA) throug wore HEVESf th ho k Y and LEVY, have been available for over 50 years (1). These authors used neutrons produced by a radium-beryllium source to determine dysprosium in Yttrium.

After the end of the second world war, the development of civil nuclear reactors allowed a large expansion for this analytical method. Analysts obtaine a conveniend t e neutroaccesth o t sn beams that became more and more intense : 10 n.cm s then 10 n.cm .s for the first-experimental French Reactor "ZOE "n 1950i , e secon5.1th 0r d fo n.cm " .s " "EL2 n 195 i d "finall 2an y 3.5.10 n.cr mfo

147 the"OSIRIS" Reactor in 1966. The concomitant progress of electronic equipmen s allowetha simplificatioa d n of the chemical separation after irradiation and has therefore spread to fields other than chemistry.

When researchers only had Geiger counters, they had to separate each radioisotope before measurement, usin methode th g f classicaso l mineral analysi year0 3 (2)d w sol no s. Scintillation detectors then allowed finer measurements of radiation and coincided with the development of column chromatography, which is well adapted to activation analysis. These new methods started with organic resin (3) and then mineral exchangers were added (A). The arriva f semiconductoo l r detectors profoundly modified the conceptio f separatioo n r activatiofo n n analysis.

Many analyse w carrieno without e sou ar d t any chemical separation and when this remains necessary, it is very selective and adapted to specific situation onl: s y certain isotopese b nee o t d isolate eliminater o d d before measurement (5).

Beginning in radiochemical laboratories, s spreaNAha A d ove year0 3 r s into many areas : suc s a h archeology, biology, environmental, geochemistry, solid physics .. .

Many laboratories, now use this method of analysis intensively and have permanent access to one or several nuclear reactors. Som f themeo e , th suc s a h Pierr e LaboratorSu e Francen i y e hig,us h power research reactors normally destine r othefo d r purposes. Others, such as the National Bureau of Standard Washingtonn i s , have thei n reactorow r n I . both cases, the reactor used is not specially destined for NAA.

II - THE ADVANTAGES OF NAA

e consistanTh surprisy t ma succesA NA e f so some. Sinc discovers it e n 1936i y , many other analysis methods have been developed som f theeo m extensively. These methods are often much simpler to apply and possibly cheaper than NAA. However, this methos ha d specific advantages which explain remaint i y swh s uniqu r certaifo e n analyses.

NAA, like other method sensitivea s i s , selective, multi-element system. Moreover it has the advantage of Nuclear Activation Methods, as it detects radioisotopes createe irradiatioth y b e atomid th f co n nuclei of the elements. This allows not only the by passing of chemical bonds and direct calibration, but also avoids most of the errors inherent in titration by classical methods.

148 Lastlys certaiha A ,NA n specific advantages over other methods of activation :

- High neutron penetration in the analysed material which simplifies calibration. e probabilitTh - e reactionth f o y d consequentlan s y the sensitivit e detectione morth th ef l o y al s i , important as the element is naturally rare (for elements with an atomic weight : A > 80). The nucleus synthesi r solaou rf o s system explains this fact D (6)(fig. .

e simultaneouTh - s analysi f severao s l samples i s possible.

All these reason stils i smose A explaith lt NA y nwh frequently used activation worlde methoth n i .d

Figure 1 : Variation of the product of natural abundance of element y theic s r neutron cross sectios a n a functio f theio n r atomic weight (6).

Ill - EXPERIMENTAL ERROR RISKS

As for other analytical methods, a large numbe f causeo r f errorso s exis NAAr fo t . Errorn sca occu t eaca r he analysi th ste f o p s show fign i n . 2

III.l. Preparatio f sampleo n d standardan s s

For sampling the causes of error are the classican i sam s a e l analysie samth e d precautionan s s e neededar e impuritieTh . s introduced during this step e irradiate,ar d simultaneously withe thosth f eo sample and will alter their determination.

149 PREPARATION SAMPLE D STANDARDAN S ?

CHEMICAL TREATMENT AFTER IRRADIATION

1MJIOACTIVITÏ MEASUREMENTS

RADIOISOTOPE IDENTIFICATIONS

CONCENTRATION DETERMINATIONS

Figure 2 : Principe of Activation Analysis

The use of a clean room is sometimes essentia o prepart l e sampleth e , especiallr fo y biological e materialsth e latterf th o r e Fo on ., principal cause f erroo s r cames froe containeth m r user irradiatiofo d n whethe mads i f o t ei r polypropylen f fuseo r do silic) (7 e a (8). These containers must be washed thoroughly before use. They must als carefulle b o y seale o avoit d d loss through vaporisation during irradiatio s wel a ns pollutio a l n due to the heating during the sealing (9).

e largTh e rang f neutrono e materialn i s s allows the analysis of samples of any shape and of a weight from one milligram to several grams. The mechanical operations and the possible resulting pollutions are thus reduced to a minimum. This is not the casr somfo ee other method f analysio s s (e.g. spark mass spectrometry), which require a special shape for the sample.

Furthermore for solid samples, it is possibl o avoit e d surface pollutio a cleanin y b n f o g the surface after irradiation. However, the cleaning is not so easy and not so efficient as can be supposed. Therefore it is always advisable to irradiate solid samples wit a cleah n surface.

e preparatioTh f standardno t no s i s particulary difficult e irradiatioth s A . n concerns only the atomic nucleus, the chemical form of the impurity doet intervenno s e durin measuremene th g f o t radioactivity oftes i t nI . direca possibl e us t o t e standardization from the natural element itself. Generally, both samples and standards are irradiated together.

This possibilit f carryino y direca t ou gt standardization absolutely indépendant of the chemical form, gives NAA a distinct advantage over other

150 analytical methods superios i t I . methodo t r s sucs ha spectrometrie f emissiono s massf ,o X-raf ,o y fluorescence achieved on solid samples, in which quantitative determinations cannot be obtained without standards of the same nature and composition as the sample analysede b o st . This requires both timd ean money. Howeve reactora n i r , sampled san standard nevee sar r irradiated exactle samth e n i y manner becaus microheterogeneite th f eo beame th .f o y principae th Thi s i s l caus f d inaccuraceo an A NA n i y it introducecentr pe ,w whatevefe erron & sa f e o r th r level of concentration. Nevertheless, it is possible to reach an accuracy of one per cent or even less, using special configuration for irradiation (10).

III.2 Irradiation

The irradiation is the greatest drawback of the activation analysis methods. This is due both to the apparatus for irradiation to be used (reactors, cyclotrons, Van de Graaff, linear accelerators...) and to the error risks which occur.

A apparatuNA n I s require particulars i d y heavy especially to have intense beams to obtain sensitive determinations thit bu ;s apparatue b n ca s used simultaneously for many other studies and the irradiations performe t increasno r NAAo d e fo d, th e running cost of a reactor.

e erroTh r o irradiatiot risk e du s e ar n avoidee b n mosn ca i d o ts know cased nan s (11).

- The Szilar-Chalmers effect and heating induce y irradiatiob d hazara e ar nd speciallr fo y fragile materials (e.g. biological samples).

e absorptioTh - f neutrono n e samplth n i se can generally be neglected except for some materials with a strong cross section (B, Cd, Ag ...). In this case, the required corrections can be carried out (12).

e recoiTh - nucleae th l o ranget r e ,du reactions, concerns only a thickness of a few nanometers with thermal neutrons and is generally inferio somo t r e micrometers with epitherma d faslan t neutrons (table II) (13).

- The abnormal diffusion under irradiation suggested by several authors (14) (15) (16), has been refute mosn i d t cases (17) (18) (19).

e otheth rn O -hand e productioth , e th f no same radioisotope starting frodiffereno tw m t elements e takeb o nt s inti , o consideratio r eacfo nh analysis.

151 4i 4l 4i c o X z < „ / in er 6 ? 2 -J § rsi 0 2 " - 6 .û u. J3 « U 0 5 — o >- ° d ; < - 5 d > Öl S ° D -û O 4. •o E - o «n u-, ~ O Q- « t— o 1» « > » 10 *\ S o _Û z Q. M -0 ^ u < 3 «n o 2 llJ - > «> o »s. £ 2 o « ° ? (A _Q u » c JD o »n _ [1. p ^ 2 0 0 r « > r *" p* ~" o es» 2 40 »n tfl 0 S o 0 «, C. X „ -û 03 < H Q O o O 8 U l tr r* - O - 6 ai er c m B '- S 6 a • _û c. T3 a> .0 N o u ": O ~* « o ^ d o w> ^ ? ° IO .c 3 _û O' "O ^ U o < „ ^f Q d ^~ o > * ? D ^J SI ~ « r ^ f~t « a; Oll •a JZ -Q a > 4J f o û. g 01 ^ L5 5 C H » ""* . o »i O S O 10 l 5 d t£ -^ U E in W -P -l O o „ X t- _ E 0 c/ re • o • O o a: - a u c c - 8 in ° O 4J 5 « 0 S ° 4J C s ° t d O 0) 3 M w u -v E •H c ^8 a. « D O •* ~* ^ •H U N ? ° ÎÂ •O _O c „ "o * - z. ° «< cc °. SS —i C « o •n 0 o 10 o rx * U) > 1-1 H L o o > s ^ ° CL ° C C U ^ 1. - rtj -^ O cv 0 H 0 > « 01 * m ° C T) r« O« *n D .a O in 'O 4* (U r-- t— 0 U ~" 3 H >8 z „ ^H c > o "ü ° m S ° •J t-> CN. 0 3 C 0 L £. o - H 0 •H C £ ~ _! « £. E O m N — "". > « 0 0 0 r^ (V o 2 ^ o m 0* ^s. 0l c 3 ra O c -H U ^ "O er c D in o >• - K K vi o o -o < 1 •H n -v' H- pi H -P -H 4-> M ?; ° S ° in « C -U o ti CJ U 4) ~i 0> Öl o •o E QJ 4J D 3s in 2 m - er QJ 4J V U-l ÖD £ 2 -i O) c (0 « 0 a o rv 0 "" O) T3 ^ IN* •n *> « (U V4 J3 t« •c. o - U) -H 0 ^ rs, i J er - u - Q) (U Z ci ^ « .c x: •n 2 ° s ° S o O f- 4J

152 TABLE II

NEUTRON * RECOINa L ENERGY 24N«* RANGE IN ENERGY

4 125 0,2

5 293 O,47

6 461 0,75

7 629 1,02

8 797 1,3

24 * For the Na produced by rann s i onle y abouu ru 3 t

reaction) T , (n generalle e ar s Th y very selective e interferinTh . g reaction o epithermat e du s l and fast neutron n onlca sy occu n elemento r s whice ar h clos atomin i e c weight.

examplen a s A , tabl I show II e blank sth s thus introduced in the determination of sodium in magnesium, aluminium and silicon. For the first two, e interferencth wels i e l know d shoulan n t givno d e inaccurate values but only a high detection limit.

TADLI II E

THE DETERMINATION OF SODIUM

APPARENT CONTFNT IN SODIUM AFTFR AN IRRADIATION NUCLEAR III THE rEACTOI ELEMENT REACTION

OSIRIS ORPHEE

23 24 Na Na(n,,) Na' 1 1

24 2 4 6 19 M9(n,D) W 3 x 10" 2 10-

27 24 4 6 Al Al(n,a) Na" 1,6 x 10" i.io-

Si 28Sl(n,p„)24Na* 3 ID'10 <1 JO"10

153 For silicon, the (n, ap) reaction is less known e threshol.Th f thio d s reactio higs i n h (15MeV), few neutrons reach this energy in a reactor and the cross sectio probabls i n y very small (22). However, whe e apparenth n t concentratio- f sodiuo n n i m electronic grade silicon is found to be several 10 g, this reaction mus takee b t n into consideration.

Uranium fission can also create the same radioisotope s thosa s e determinatioeth user fo d f o n some elements wit atomin ha c number d 140arounan .5 9 d Tabl showV I e s this risk whic wels i h le th know r fo n determination f rarso e earth geologican i s l samples (23).

TABLV I E

INTERFERENCE F URANIUE O SDETERMINATIO TH N I M F SOMO N E ELEMENTS BY NEUT1ON ACTIVATION ANALYSIS

APPARENT WEICHT lug) E CLEMENTH F O T PRODUCED ELEMENT RADIOISOTOPE YIED FOR 235U FISSION F URANIUO g u B1 MY WITH THER'VlL NEUTRONS USED TESTED Calculated Determined

Zr 95Zr* 6,4 % 10,5 9,2

Mo "MO* 6,1 1 3,3 2,3

La 140B," 6,3» 0,002 O.OO3

V.140

Ce U1ce 6 % 0,3 0,2

Values are calculated for an irradiation of 72 h In a flux of 1.1O n cm" .s" . th Determinatio e performenar dOSIRIe afteth n n IrradiatioSi a r pooh l2 7 reactor f o n .

III.3. Chemical treatment after irradiation

A chemical and eventually a mechanical treatmen generalls i t e irradiateyth carrien o t dou d sample befor e radioactivitth e y measurements, except for the analyses using very short-lived radioisotopes.

For solid samples, this treatment usually includes the cleaning of the surface, which theorically allow avoio st d surface contamination introduced during sampling, conservatiod an n irradiation. Method activatioy b s n have thi vera s y great advantage over other methods obtaio .T bese th nt resulhoweves i t i t r necessar e particularlb o t y y careful, speciall ultra-tracr fo y e determination 1s. ( 1 )

154 Thanks to the progress of electronics and computers, it is now possible to determine a large numbe f elemento r wida n i es rang f concentrationo e s withou y chemicaan t l separation. This avoids difficultie d erroan s r risks, speciall materialr fo y s that are difficult to dissolve (e.g. alumina) and for volatil elements that are difficult to determine without losses (e.g. mercury).

However in some cases, chemical separation prior to the measuring of radioactivity remains advisabl d ofteean n indispensabl o obtait e e th n sensitivity required. The methods of separation used derived from classical methods wit preferenca h r fo e chromatography over columns and liquid-liquid extractions n comparisoI . n with other methodsf o e ,us radioisotope maio tw n s advantagesha : s

e suppressioTh f blanke stago nth f analysit o esa s allows the use of a wide range of reactives, without the long and dreary task of ultimate purification and without compromising of the sensitivity obtained. f inactivo e us e Th carriers minimize e lossesth d an s allows their measuremen r eacfo th separation.

An analyst using activation techniques should never make an incorrect radiochemical separation, because he can control the radioactivities at each step of the separation.

III.4. Radioactivity measurements

The performances of NAA have been much improved by technical progress in measurement systems. w commono s o determinI t n o fortt p u ey elementa n i s concentration range unde e p.p.m th rd sometime an . s r undesevera o e p.p.b.e th ron ln i ,sample s irradiated simultaneously.

e radioactivitTh y measurement vere ar sy convenient to automatize. This automatization requires rigorous testing, but afterwards, it is possible to analyse a large number of similar samples with reduced labor costs.

The present evolution of the computer systems considerably reduces the cost of measurement apparatus e associatioTh . multichannea f o n l buffer with a microcomputer such as an IBM-PC, reduces the e acquisition-treatmenth cos f o t t syste a facto y b m 3 r or 4 in comparison with the former systems, such as TRACOR 4000, ORTEC 7000 or NUCLEAR DATA Cosynus.

Moreover e softwarth , e eth user fo d mathematical treatment of gamma spectrums with a personal microcomputer, are more accessible for the analyst than those written in assembly language. Thus

155 possibls ii t o havt e a ebette r adaptatioe th o t n sample analyseb a betteo t ed an dr e controth r fo l results obtained.

The nuclear characteristics of the radioisotope sA (half-time NA use n i d , typf eo emission, gamm energiey ara s ... generalle )ar y well known (24) (25). When sample standardd san e ar s irradiated simultaneously, incertitude neutroe th n o sn flux and on the values of the cross-section do not intervene.

Consequently, the determination of the concentration is very easy: the radioactivity measured at the same time after irradiation is directly proportional to the number of atoms of the same elemen samplen n standardsi ti d san .

However, some radioisotopes have very similar nuclear characteristic confusede b n ca d .san In fact, this error risk is exceptional when the measurements are correctly performed : both gamma-ray energies and half-life must be controlled. Nevertheless this risk cannot be entirely neglected.

TABLE V

DETERMINATION OF TITANIUM BV THE NUCLEAR REACTION

OTHFD AN Tl(n,pR * c INTERFERINS ) G NUCLEAR REACTIONS

APPARENT CONCENTRATION NUCLEAR PRODUCED BY THE ELEMENT ELEMENT Calculated Determined

Ti 47Tl(,,.p)«7Sc* 1 1

2 V 50V(n..)47Sc« x 10"3 2 x 10~5

46 1 1 C» Ca(n..()<" 2 X JO" x lo"3 8 C- a x ID" 4 (sou 3: )d C s

<7SC*

8 199 2 2 rt " Ptln,,) Pt 10 x 8 6. 5 x 10 ( sous Cd : 4 x 101) \ '" 199Au"

Values are calculated for an Irradiation of 72 h In a flux of f 1.1o d On an ..c s ~. ms . m .c , 1.1 n O

OFTTFR1INATIONS ARE PERFORMED AFTER A 72 h IN THE OSIHIS POOL REACTOR

Tl/2 • 3,42 days Tl/2 - 3,2 days

E - 159 KeV 8 K«15 V - Er

156 For example in the determination of titanium using the 47 * 199 * 199 * s ha u t A producF decaß y e f b to yth , c S e samquity energ th e samra e th d ex ehalf-lif an y s a e

n thic I S (Tabl s necessars . i case V) et th i e o 4t y7 * e higus h resolution spectrometry measurementa r o s radiochemical separation after irradiation.

IDEVELOPMEN- VD PROSPECT AN E FUTUR A TH NA R EF SFO TO

A large number of papers presented in June 86 during the last meeting entitled : "7th Modern Trend n Activatioi s n Analysis" (26), show that no t onlA continueNA y usee b do t sthroughou worle t th t bu d also thas applicationit t e developinar s severan i g l domain d particularlan s e threth n ei yfollowing .

1. Ultra-Traces analysis

In this case the suppression of pollution after irradiation and the showing of the losses during the chemical treatment, allow to reach theoretical limits in the analysis of real samples.

Table VI shows the results actually obtaine e analysith n i df hig o s h purity silicon. Generally the detection limits obtained on this material are better than those calculated previously in Table I. Several hundreds of silicon samples are analysed each year in the "Pierre Sue" Laboratory, at a very competitive cost n comparisoi s n with other methods, such as : S.I.M.S., I.C.P., S.M.S., ... which e mucar h less sensitiv practicn i ee th r efo thaA NA n detectio mosf o n t metallic impurities.

2. Correlation studies

Here, the multi-elementarity of the NAA allows to compare quantitatively the concentration of a large number of elements in a same sample and in a large rang f concentrationeo . This applicatios i n widesprea archeologyn i d , biology, geochemistrd an y environmental studies.

For example, in archeology the multielemental analysis of clay allows to trace the origin potterf o s y (27) thin .I s case instrumentath e l NAA competes with X-ray fluorescence. The INAA allows the determination of a greater number of elements with greater accuracy than X-ray fluorescence, which is a complementary method to determine some elements ( Si, tha) difficule Pb ar t d an o P t , Nb , Zr , Y , Ti , Ca measure with the sensitivity required by NAA (28).

157 TABLE VI : Results obtained for high purity silicon by NAA (sample e weighinth n g irradiatei I h : g 72 r fo d OSIRIS reactor at C .E.N./SACI.AY)

CONCENTRATION IN »g. g' (lu' 9)

ELWFNT NUCLIDE HALF-LIFE Detection Detected Limit

1 10 -2 z *l *S 250 a 0 1 1 < 8.,o" 76 -3 _ 3 Ai As 16,3 h < I.IO I.IO IS8 4 . t Au Au 2.7 a 6,6.10" 1 .10 131, Bi &3 11,8 a < 1.5 1.5 62, 2. 2 - Br ir 35,3 h < 1. 10 I.IO ) Ct "5Cd 44,6 a -: 3 10 3,0"' I4 1 - 1 Ce 'c. 32,5 a < I..O 1 . IO

Co '°Co 5.27 y < I.IO"1 LID'1

2 2 Cr MCr 27.7 d 4.5.10 2.10- I34 . 2 - 2 C. c. 2.O6 y < 1.10 1 .10 . 1 Cj "cu 12,7 h < I.IO 1 . lo" 1 . 3 - 3 Eu '"EU 13,4 y < 2.10 2. IO s Tt 'r. 44,5 a i < 4 2 2 2 - C< ' c. 14,1 h < 5. 10 5 10

I8l 3 11f «f 42,4 a < S.lo" S10-

. 1 _ 1 203 46,6 d < 7.10 7.10 "S HB T In '""in 49,5 d < 3 10 ! .0"'

4 _ i. ? 2.10 2.10 Ir " Ir '4.0 d 1 1 - K ", 12.4 h <. 6 10 6.10 . 4 _ 4 L» "°l. h : . « 0 1 . 7 < 7 10 4 it Lu "7Lu 6.' d < 6.10 6 io" 2 2 < 1 10 l.io Io "Mo 66,0 h [ h J 'S. i:.oi n < 310" 3 ID''

N. 58Cc 7' ,3 d j <; 3 199 Pt Au 3,2 d oi ". s < 03 I < 1.10 I io" Pu • RJ 39.3 d 2 '0 ' Su 'Nb 6« . 2 c 4 10 j A8 2 I«'3 <;, ÎC 6! f d <• 2.lo" 1 75 , .0"' S« s, 1 19.d 8 < 110 4 153. 4 ,0"' Sm am 46.8 h < 410 4 Sr 89Sr 50,6 d < 4 2 . ! lo"2 T. 182T. 115,4 d < I.IO , . 3 J n> "°Tb 72, 1 d < 6.10 6 10 . 1 2 10" ' 1c 1:'l 8,0 d o i : ••

4 30 Ti 'sc 3.'. d < 30 3 3 2J3 l.S.lo" Th.. P. 2'.0 d < I.S.IO" , 2 23 2.4 < 210" 2 ,0" U '.-P d ; u I87v 23,9 h 5..0-' - o : l 3 ^ Tb I7;Y, »,2 d < ; . 10 " 0 1 1 i , Zn "Zo 243.9 d < 2 10 2. 1C

2 Zr "zr 64 d 2 <

158 e advantagbiologn i Th o t A s NA i y f o e determine very low concentrations in very small samples (biopsies on babies and infants). Table VII summarizes these applications (29).

TABLE VII

Importance of elements in various areas of the Sciences of Life and their detection by NAA from (29) (Indication of the detection level : 1 < l.p.b. ; = 21.1 0 p.p.b1O-1O: 3 O ; p.p.b..O 1O p.p.b> 4 ) ; .

Importance tn Mode of analysis Leve f detectioo l n radiochemie th f o - Nutrition Environment!) Instrumental Radio chemical cally isolated Element pollution froup separation nidioisolope

AI Ai

Ca Cd O Co O Oi F ft Ht I K Li Ml Mn Mo N» Ni P Pb Rb St> Se Si 5n Sr Tl V Zn

e verTh y hige h th sensitivit n i A NA f o y determination of rare-earths is used in geology to modeliz larga e e numbe geochemicaf o r l processes (30).

3. Standardization applications

oftes i A n NA use e o certift dTh y standard samples. Moreover, radioisotopes produced by neutron activation can be used to show the pollution and loss risks in chemical separation of classical methods.

NAA, like other methods, cannot avoil al d causes of errors. However, the radioisotopes allow to eliminate or at least to control most of these causes. In intercomparison runs to establish standards, the activation method e thosar s e which generally achieve the values finally retaine mose th dt quickly.

Great organizations like N.B.Se th th e n i . U.S. daily use NAn to certify their standards. The accuracy of this method proves to be superior to the

159 accuracy imposee dispersioth y b d n betwee e resultth n s obtained by the other methods.

Developmen différentn i A NA f o ts fields

There are three principal fields in which the uses of NAA are spreading : biology, geology and environmental studies. Table VIII.

TABLE VIII

NUMBER OF PAPERS GIVEN DURING THE 7th

MODERN TREND N ACTIVATIOI S N ANALYSIS

BIOLOGY 42

GEOLOG 6 D COSMOLOG3 AN Y Y

ENVIRONMENTA5 L3 STUDIES

METALS AND SEMI-CONDUCTORS 14

ARCHEOLOGY 5

FORENSIC APPLICATIONS 2

e futureth n I , application n biologi s d an y environmental fields are likely to continue increasing. For these applications, the response time methoe th sometimes f i do f o se us delayed e th o s , short-lived radioisotopes should be extended. In geology, the studies using rare earth determinations are widespread, but the studies of the elements of the platinum family could be developed. For metals and semi-conductors uses i dA mostl,NA r qualitfo y y controls and for calibration of measurements of physical properties. After calibration, these measurements can be used for routine controls. As industr mor s d i mory an e e intereste n higi d h purity materials, there n increasewila e b l d demanr fo d highly diversified standards n otheI . r fields, sucs a h archeology and criminal investigations, NAA is not wide spread, probably because potential users do not have sufficient training.

The main drawback to a greater and quicker developement of NAA is that access to nuclear reactors is difficult e latteth e speadins A ar .w r ne o t g countries A ofte,NA n develop n relatioi s o t n theeconomic needs in these countries.

Another drawback is probably psychological. Many users consider that these techniques are very heavy, expensive and even dangerous n factI . , initial investment e highar st bu , late e e methoaccese costh th rth f o t eass o d i t ds an y

160 o highen NAs i Ar thae cos th nf otheo t r modern analytical methods which claim similar performances. The risks of irradiation and contaminations exist, but accident more ar se hande thaon n,e rareth n .O radioanalysts are aware of the danger and are fully equipped for detection and protection. On the other hand, the level of radioactivity required for sensitive determination remains very low.

V - CONCLUSION

Althoug s fifti t yi h years old, neutron activation analysi stils i s l widely used continuean d s to develo l ove worlde o otheo nowal pn th t r w , p ne rU . analytical methods have been able to replace it for some applications. The use of active isotopes, directly produced from nucle f elemento i e b o t s determined, gives it exceptionnal properties : sensitivity, selectivity, accuracy and the capacity to determine simultaneousl a ylarg e numbe f elemento r n i s a large rang f concentrationso e e nee havo Th .t d e acces nucleaa o t s r reactor constitute mais it sn drawback. As long as this access is further promoted and generalized A application,NA s will continuo t e develop.

REFERENCES

HEVES- 1 K.DANS- , LEV YG. . IH K ACAD, _L4, 19365 ,

2 - ALBERT P. - Annales de Chimie, 1_3, 1955, 827

3 - AUBOIN G., DIEBOLT J., JUNOD E., LAVERLOGNERE J. Proceedings of Modern Trends in Activation Activatio Colleg- n e Station, Texa4 s34 196- 5

GIRARD- 4 PIETR, IF. SABBION, AR. . IE J. Radioanal. Chem. 5, 1970, 147

REVE- , FEDOROF5 LG. Nuclea- . FM r Instrumentd san Method 143, s . 7 197727 ,

BURBIDG- "Synthesi- 6 . al Ed E.Mf elementan o s. n i s Stars" Rev. Mod. Phys., 29, 1957, 547

HEYDOR- 7 , DAMSGAARNK. Talanta- . E D , 198229 ,, 1019

CORNELL- 8 HOST, VERSIEC, R. S EJ. . J K Talanta, 2^, 1982, 1029

MAZIER- 9 GAUDR, EB. GRO , COMA, YA. J. S . D R Radiochem. Radioanal. Letters, 28, 1977, 155

161 1DUGAI- 0 Proceeding- . NF e 196th 8f so International Conference Modern Trend Activation i s n Analysis. Gaithersburg (USA) 7-11 Oct.,19684 47 . ,p 11 - REVEL G. - Analusis 12, 1984, 506

12 - FEDOROFF M., LOOS-NESKOVIC C., REVEL G. Anal. Chem., 5J., 1979, 1350

13 - DELMAS R., FEDOROFF M., REVEL G. J. Radioanal. Chem., 79, 1983, 265

1ALBER- 4 , BLOURTP. CLEYRERGU, IJ. DESCHAMP, EC. . SN LE HERICY J. - J. Radioanal. Chem., 1., 1968, 297, 1 43 d an 9 38

1DELMA- 5 , FEDOROFSR. REVE, FM. . LG Radiochem. Radioanal. Lett, (30), 19779 ,20

Radioanal. J 1 NIES- 6- . ES . Chem. (72), 1982,9

17 - DELMAS R. - Thèse de Doctorat d'Etat, Paris 1982

18 - DELMAS R., FEDOROFF M., REVEL G. Anal. Chim. Acta 120. (1980), 227

19 - DELMAS R., FEDOROFF M., REVEL G. Trac Micropobd ean e Technique , 19842 s 7 6 ,

REVE- , LESBAT0 2 LG. ALBER, SA. . TP C.R. Acad. Se. Paris, 258. 1964, 148

21 - LESBATS A. - Thèse Paris 1968 Ann. Chem., 4, 1969, 34

22 - NIESE S. - J. Radioanal. Chem., 38, 1977, 37

MEYE- Radiochem- 3 . 2 G R . Radioanal. Letters, 52,19823 23 ,

24 - ERDTMANN G., SOYKA W. - The gamma rays of the radionuclides - Editor Verlag Chemie - Weinheim - New-York 1979

25 - ERDTMANN G. - Neutron Activation Tables - Editor Verlag Chemie - Weinheim - New York 1976

26 - 7th Modern Trends in Activation Analysis 1986 June 23-2 Copenhage- 7 n (Denmarke b o T )- published in the J. Radioanal, and Nucl.Chem.

PERLMA- 7 2 , ASARArcheometry- I. N . F O , 1969U_ , ,1 2

SABI- Thès- 8 . 2 RA e Paris 1986

29 - BRATTER P. - Radiochimica Acta, 3_4, 1983, 85

30 - TREUIL M., JORON J.L., JAFFRREZIC H. J. Radioanal. Chem., 71, 1982, 235

162 COMPARISO NEUTROV Me 4 1 F NNO ACTIVATION ANALYSI COMPETITIVD SAN E METHODR SFO DETERMINATIO OXYGENF NO , NITROGEN, SILICON, FLUORINE AND OTHER ELEMENTS*

R.W. BILD Sandia National Laboratories, Albuquerque, New Mexico, United States of America

Abstract

14 Mev neutron activation analysis (14 MeV NAA) makes use of small particle accelerators to produce 14 MeV neutrons from the D-T reaction. The neutrons produce radioactive isotopes in samples by the reactions (n,p), (n,2n) and (n,a). Gamma rays emitted are counted to determine the amount of the target element present. Major applications have been determinatio f totao n , 0 l N or Si in solid and liquid matrices, but the technique can also be applied to determine concentrations of about 30 other elements including F, Cl, AI, P and Mg. Detection limits are a few micro- e besgramth tn i scase d milligraman s r mosfo s t otherse Th . method has advantages of being nondestructive, fast and insensi- tiv o samplt e e inhomogeneities t lendI . s itself especially well to sequential analysi e samth e f o sampls y severab e l techniques ano samplet d s thae difficular t o dissolvet t . Portable generators have been applied to industrial situations and to well logging. Major disadvantages are the necessity to house a radiation producing instrument e equipmen e th lace cos th f th f ,o ko td an t useful neutron reactions for some important elements. Accuracy (typicall? relative10 o t * d precisiorelative1 5 ± an y) o t 1 (± ?n are comparable to competing techniques. For determination of low levels of 0 and N in most metals inert gas fusion is more rapid and sensitive; elemental analyzer is more sensitive for 0 and N in organics t chemicaWe . l methods rarely havy advantagan e e over r solifo A d NA sample V 1Me 4 s when concentratione th n i e ar s detectio A methodsNA V n Me limi .4 1 Futurte rangth ef o e develop- e fielmentth n di s wil e larea th com f simplern o si e , more portabl d highean e r neutron output generator designs.

INTRODUCTION

V neutro1Me 4 n activation analysiV NAAMe 4 ) becam(1 s e practical wite developmenth h f relativelo t y inexpensivew lo , beam current, deuteron accelerators in the late 1950's and early 19_60's. Outputs of 14 MeV neutrons were originally about 10 t sincbu e thes n have increased several order f magnitudo s e

* Work performed at Sandia National Laboratories and supported by the US Department of Energ\ under contract DE-AC04-76DP00789

163 The generators are well distributed in laboratories around the world, both in developed and developing countries. Applications have been primarily determination of bulk oxygen, silicon and nitrogen content a wid f eo s variet f soli o yd liqui an d d samples. Applications to many other elements are possible and regularly reported. The intent of this paper is to review the technique, discus s currenit s t application d assesan se futurth s e rold an e possible future developments of the method. This paper is not intende o providt d a ecomplet e bibliograph e l paperth al n f o so y topic; more complete referenc ee foun b list [1,2,3,'On y i d ma s - Competing analyticaA l NA method V e majo Me th r eac4 1 fo rsf o h applications wil e describeb l d discussean d d with emphasie th n o s strength d weaknessean s f eaco s h method.

DESCRIPTION OF METHOD

Principles 14 MeV neutrons are a product of the deuterium-tritium (D-T) fusion reaction: 17.+ n 1 6+ MeVe H 4 . > —— H 3 + H 2 e Approximatelenergth f o y V releaseMe e 4 th 1 y s carrie y i db f of d neutron as kinetic energy. Small deuteron accelerators specifi- cally designed for 14 MeV neutron production are generally used, though other types of particle accelerators, such as van de Graaff accelerators n als ca ,e used b on eithe I . r cas a beae f o m deuteron s acceleratei s d int a otritiu m loaded targete Th . sample irradiateb o t e s positionei d d e targeclosth o t et while d e thegeneratoan th nn o move s i rd (usuall a pneumati a vi y c transport system a gamm y o detecto)t ra a r countingfo r . Excellent detailed discussions on the theory and practice of generator design and operation may be found in [2,3,5,6],

Equipment Required Commercial generators are of two main designs. Open tube types typically accelerate a pure deuterium ion beam into a tritium loaded target e neutroTh . n output decreases with target use, with targets usually being usefu w operatinr jusfe fo la t g hours. Target replacement requires only a few minutes, but several hours of pumping may be required to restore the vacuum to e optimuth m level A typica. l targe6 Ci/c1. o mtt loadin8 0. s i g (3 x 101° to 6 x 10 1° Bq/cm2) of tritium. Sealed tube neutron generators have the entire ion source- accelerator-target assembly permanently sealed at the factory. A mixed bea f deuteriuo m d tritiuan m s acceleratei m d onte th o target, simultaneously producing neutrons and regenerating the tritiu me targetleveth n i l . Useful r mortubo 0 e 10 lif s i e hours, after which the entire sealed tube is returned to the manufacture r replacementfo r . With this typ f generatoo e e th r operator never handle n opea s n tritium source. The commercial generators typically operate with accelerating potentials in the 150 to 200 keV range and at 1 to 5 mA target currents. Neutron output is around 10 s

164 e Graafd Somd n othean va fe r particle accelerators than ca t be used with deuterons can also serve as 14 MeV neutron sources. Neutron output will depend on the target and beam currents and on individual design factors.

In addition to the generator, a room or cave with suitable personnel shieldin r neutronsfo g , gamma rayd othean s r radiation produced by the operating generator is required. A pneumatic sample transport syste s usualli m y installe o carrt d y samples between the generator and the counting station. The counting equipment is located in an adjacent room with the generator controle operatorth d an s .

Counting equipment used for 14 MeV neutron activation is the typical sort foun n radioisotopi d e counting labs:d Nalan e G , Ge(Li) detectors wite appropriatth h e shields, power supplies, amplifiers, multichannel analyzers and data reduction computers. r oxygefo A nNA requireV 1Me 4 l detectorssNa onle th y ; many other A applicationNA V 1Me 4 s work better wite higheth h r resolutiof o n modern, high efficienc r Ge(Lio e G y ) detectors.

Cos f Equipmeno t t Settin a complet V p neutrou gMe 4 1 e n activation analysis facility with all new equipment can be expensive. Estimates of current prices, all in US$, are:

V neutro1Me 4 n generator $225,000 Pneumatic transport system 40,000 Detectors [Nal and Ge(Li) with shielding] 30,000 Electronics for counting equipment 10,000 Multichannel analyze d computean r r 20,000

Total $325,000

Cost r buildinfo s r remodelino go safelt b la y a housg d an e shield the generator are not included in the above. Expenses in setting up a 1 4 MeV NAA capability could be significantly less if ther s existini e g counting equipmen r eveo tn existina n g particle accelerator which could be used.

Reaction sV Neutron Me wit 4 1 h s The most commonly used 14 MeV activation reactions are (n,p), (n,ct d (n,2nan ) ) reactions. Cross sectionr thesfo ) e (o s e millibarreactionth n i e nar s (mb) range, generall a factoy f o r o 100t 0 010 lower than thos r capturfo e e (n,Y) reactions user fo d thermal neutron activation analysis. Capture reaction cross sections for 14 MeV neutrons are usually too small for the reactions to be useful. Some of the more commonly used 14 MeV neutron reactions are listed below (data from [7,8,9]). Potentially useful reactions with many other elements also occur ([2,7,10]).

l6 l6 0(n,p) N t1/2 = 7.14 s o = 39 mb

1%(n,2n)13N t = 9.97 min o = 7.0 mb

165 •*j Q O ft b m 0 23 = o Si(n,p l n A )mi t3 1/2. 2=

19 l8 F(n,2n) F ^/2 = 109.8 min o = 55 mb

Gamma rays produce e reactioth y b d n product e countear s d usean d d to calculat concentratiothe e elemenof n t present.

When selectin a reactiog r use,fo n possible interferences must be considered. For example, the reaction

19F(n,a)1°N t = 7.14 s o = 20 mb

produces the same product isotope as the reaction used to determine oxygen. Thus small amounts of oxygen cannot be determined in the presence of an excess of fluorine. Other interfering reactions can occur which produce a different isotope than the f interesto e thae on on tt emitbu , s gamm e asam th r rayo e f o s nearlf interesto e same th yon e n these I energ.th es a ycase s e differinth g o half-liveisotopetw e n ofteth ca s e f useb o nso t d resolve the interference.

Advantage f Activatioo s n Analysis The advantages of the 14 MeV activation analysis method are its nondestructive nature, speed, freedom from contamination and, for some element/matrix combinations, a lack of other good analytical methodse b A nee .o NA t d Sample v e analyzeMe b 4 o 1 t s y b d cut to fit the irradiation container being used, cleaned (where appropriate d weighe t otherwisan ) bu , dg) (typicall0 e1 o t 1 O. y the metho s nondestructivei d . This eliminates problem f samplo s e dissolution, reagent contamination and loss of volatiles. There o idilutio n se sampleo dissolutionth t e f o ndu s a rela d an -; tively large volume of the sample may be included in the analysis, making the results more representative of the total sample. The method also determines the total amount of the element present independent of chemical state. This can be an advantage n thai , t o problethern s i e m about unexpected insoluble or unreactive phases, or a disadvantage, in that it is impossible t chemicatge o l speciation data.

Because of the relatively low neutron fluences used in 14 w reactiolo e th nd crosan A MesNA V sections, ther s veri e y little residual radioactivity left in the sample after a day or two of decay. The same sample may be easily used with other analytical techniques after 14 MeV NAA is completed. This feature is especially important for small, rare or inhomogeneous samples. e besth t On f o examplee r this sequentiafo s wa s l analysi f somo s e lunar material [11]. A sample was first analyzed by 14 MeV NAA for oxygen and silicon. The same sample was then irradiated in a reactor for analysis by thermal and epithermal instrumental neutron activation analysis (INAA) for an additional 20 to 3- elements. After completio f INAe o sampln th As re-irradiatewa e d r analysifo r severafo s l additional element y radiochenioab s l neutron activation analysis. In all, concentrations of about 50 elements were quantitatively determined in the same <1 g sarr.ple.

166 Analysis by mass spectrometry, atomic absorption spectroscopy, optical petrography, electron microprobe, or other techniques could equally well have followed the 14 MeV work. This provided a large coherent set of data where inter-element trends could be examined independentl e homogeneitth f e yo samples th f o y . Disadvantages of 14 MeV Neutron Activation Analysis The main disadvantage of 14 MeV NAA is the extra precautions neede o safelt d y handl e radioactivitth e y produced e samTh .e type f administrativo s e control d personnean s l dosimetry required of workers in radioisotope laboratories are needed for a 14 MeV NAA lab [12]. In addition, personal dosimetry for neutron exposure shoul e providedb d . When operating e generatoth , r itsel a stron s i f g sourcf o e gamma radiation as well as neutrons. This requires that the generator be housed in a specially shielded room (cave) or under- ground pit. Dependin e specifith n o g c generatore outputh d an t size of the area shielded, several meters of concrete or earth coul e requiredb d . Boron r lithium-loadeo - d materiale b n ca s use n criticai d l areas suc s doora h o obtait s n necessary shielding with less bulk. The design and construction of a room to V neutroMe hous4 1 na e generato n cosca ra tsignifcan t fraction e generato th e cos th f o t f o r itselfV e advantagMe On 4 .1 a f eo generator as a neutron source is that when the generator is off, neutron production stops completely and gamma radiation quickly falls to a very low level. This makes the equipment safe and easy to work around when it is not operating.

Generators with removable targets present some hazarf o d tritium contamination for equipment and personnel. Target changing e donneedb o et s wite samth h e s require i car s a e n i d handling any radioactive sample in order to prevent this contamination. Large amountn io f tritiue o s th n y buili mma p u d pump that maintain e higth s h e generatorvacuuth n i m . Care needs e use b o contai t o dt n this tritiue generatoth f i m s movei r r o d decommissioned. Sealed tube generators have all tritium contained in the sealed section of the generator and thus have a much lower tritium contamination hazard. In all cases the generator room should be properly ventilated. e sampleTh s analyzed become radioactive after irradiation, but because of the low neutron flux used and the low cross section for 14 MeV reactions, their maximum activity is very low compared to neutron-activated samples from a reactor irradiation. The primary 14 MeV produced isotopes also have short half-lives. Most samples will decay to close to background radioactivity levels after a day or two. Thus sample handling presents no significant problem.

Anothe s thai r A t disadvantagNA som V eMe importan4 1 f o e t elements cannot be determined because of the lack of suitable fast neutron reactions. Som S f thes o e, Be e , C element , H e ar s and Bi.

167 MAJOR APPLICATIONS OF 11 MEV NEUTRON ACTIVATION ANALYSIS AND COMPARISON TO OTHER METHODS e mosTh t common application V neutroMe 1 1 nf o sactivatio n analysis have been determination of oxygen, nitrogen and silicon. These application e discussear s d individually below. Some other applications are discussed in a later section of this paper.

Oxygen The determination of oxygen concentrations has been by far the most popular application of 11 MeV NAA. The reaction used is:

l60(n,p)l6N t = 7.11 s Gamma rays: 6.128 MeV (69.0?) V (5.0?Me 7.117)

Escape peak 6.60: at V s 6Me 6.09V 5Me 5.617 MeV 5.106 MeV

The energie f theso s e gamma raye highear s r than essentialll al y other 11 MeV NAA produced gamma rays except those fron B(n,p) Be. By setting a lower level discriminator (LLD) on e countinth g equipmen o recort t d only those gamma rays with energies above 1.5 to 5.0 MeV, interfering lower energy gamma rays are excluded, leaving only N gamma rays to be counted. (An exception to this is discussed below.)

e majoTh r direct interference wite determinatioth h f o n oxygen concentration is fluorine which produces the same N isotope reactionth y b e :

. 19mb F(n,a) 0 2 = a l6N Thus oxygen cannot be determined in the presence of an excess of fluorine. A large excesss of boron can also interfere with the high energy gamma ray measurements, otherwise the oxygen concentration, if at or above the detection limit, may be determined in essentiall y matrixan y . In some samples the matrix element activates and produces an isotope tha tn abundana give f of s tf relativelo gamm y ra a y high energy t stil- bu ,Si D lsettinga e cas LL belof th o e e n I th w . rich matrix (e.g., SiC, Si(Li) alloy) the reaction is pQ pO t 1.7a a gamm Si(n,py d 8 = ra MeVa2.2an l witn A 8. )mi t h op When a large amount of Si is present, the Al activity is very high. Random triple coincidence of the 1.78 MeV gammas occur in e detectoth r producin n apnarena g 3 tMeV 5. pea t ,a k well abo/e the LLD setting. This interference is usually significant only at hig i concentrationS h s (>3 0o 10?)t . Sinc e magnitudth e f o e the interferenc i contenteS increasee th e cub th f ,o e s whea s n e interferencth e vers b presenti e n y ca sever t i , e [13,11,15].

168 e interferencTh e overcomb y ma ey mathematicab e l correction following repeated counts [15] or by using faster electronics. e efficiencTh e higf N Ge(Lith o y h t a )w detectorlo o to s i s gamma ray energies to be of general use to resolve the gamma ray lines froe interferinth m g activity. The detection limit is about 10 pg for oxygen. For typical system designs and sample sizes, this corresponds to detection e loweth t rA level s necessar. i ug t i s0 10 o t yo t limit 1 f o s replace air in the irradiation vials with an oxygen-free gas such as nitrogen, to use low oxygen sample carriers and to run blanks. r hourpe 5 .1 o t 2 1 e rat th f o et a n ru Sample e b y ma s Some recent examples of the application of 14 MeV NAA to the determination of oxygen are:

1. Reagent chemicals [16]. . 2 Total oxygen [17,18 d organi]an c oxyge coan i n l [19]. 3. Coal conversion liquid [20]. e powderF . 4 s prepare y plasmb d t [21]je a . 5. SioNh and Sic fums on Mo plates [22]. 6. Mg powders and ingot [23]. . 7 Synthetic metal nitrides [24]. 8. Rocks [25]. 9. Nb alloys [26].

Competing Methods for the Determination of Oxygen Several other techniques are used to determine oxygen in some materials. Eac f theso h e technique s usualli s y usefua r fo l specific type of sample. Inert gas fusion and vacuum fusion are two similar methods for determining oxygen, usually in metals. Inert gas fusion has become the more widely used of the two techniques and will be discussed further. Equipment manufactured by the LECO Corporatio s typicai n f curreno l t iner s fusioga t n instruments. A sample is placed in a graphite crucible and heated to about 1700° o mele t samplC th td frean e e dissolved t oxygenho e Th . graphite crucible reduces oxygen-containing compounds and frees oxygen as CO or CO . All of the freed gases are swept away from the crucible and through a hot CuO column by a flow of He gas, converting the CO to CO and driving the reduction reaction to completion. The CO gas is purified by passing through additional columns to remove water, etc. and then to a detector (e.g., infrared) to quantify the amount of CO produced. [27]

Inert gas fusion determination of oxygen works well for steels and many other metals. It is not suitable for very volatile metals (Na etc., ,Li ) becaus f condensatioo e e metath lf o n vapor in cool parts of the gas train. Some metals also have melting points too high to allow easy analysis by this method. All form f oxyge o st graphit ho e converteny b b tha n O eC ca t o t d e detectedar . This includes oxides foun n mosi d t common samples,

169 e trub r som fo et eno oxidey buma t s foun morn i d e exotic materials. Total analysis time is several minutes per sample (after sample cutting and surface cleaning). Sample size is typically about a gram and detection limits are less than 1 pg/g. Equip- ment costs are about US$ 55,000. Inert gas fusion is more sensitive, faster and about equally precise as 14 MeV NAA for oxygen determination in most metals. The inert gas fusion technique is destructive. Elemental analyzers (manufacture y Perkib d n Elmer, LECO, Control Equipment Corp d othersan . e use )o analyzar t d r fo e oxygen, nitrogen, hydrogen and carbon in combustible materials, usually polymers and organics. Fluorine, phosphorus, silicon and many metals interfere with the analysis. In the oxygen method, the sampl s pyrolyzei e n heliui d m over platinized carbon. Oxygen in the sample is converted to CO and then passed over hot CuO to convert CO to CO After some purificatios i n O stepsC e th , 2' detected and quantified using a thermal conductivity or other detector. The pyrolysis method works well for organics and can be applie o somt d e other sample f cari ss take i e o insurt n e that all oxygea know r e oxygenn(o n th par f o s freedti ) . Sample size is about 3 mg which can lead to inaccuracies if samples are not homogeneous at this scale. Detection limits are about 0.5 to 1 . w% tAfte r equipment calibration (which take n hou a sr two) o r , e rat f th r abouhouro e pe t a 3 t. n Equipmenru samplee b n ca st costs are approximately US$ 40,000. 14 MeV NAA is faster, better for inhomogeneous samples, about equally precis d sensitivan e s a e an elemental analyze r determinatiofo r f oxygeo n n concentrations. e elementaTh l analyze e betteth s i rr metho r smalfo d l (<0.) g 1 samples but is destructive. Several chemical methods for determining the amounts of oxygen-containing phase metaln i s e describear s n [28]i d . Most of these rely on dissolving the matrix metal but not the oxides. The oxides are collected and weighed or determined by some other chemical means. The oxides collected are very dependent on dissolution conditions and can be tailored to give information on oxygen speciation. The chemical methods are most useful where only a few samples need to be run, when access to faster techniques (14 MeV NAA, LECO, etc.) is not available or where the speciation informatio s importanti n . Equipment require s thai d t founa typica n i d l chemistr b plula ya scontrolle d atmosphere chamber or glove box. Analysis time per sample is generally several hours.

Nitrogen Nitrogen is determined by 14 MeV NAA using the reaction:

1 13 Vn,2n) N t /2 - 9.97 min. Gamma emission: none Positron emitter: V annihilatio0.51Me 1 n radiation (199.6$)

e relativelTh y long half-lif requireN r fo es (and permits) long irradiations (1 - 10 min) and counting times (2 - 20 min) compared to oxygen determinations. In order to resolve the 0.511

170 MeV annihilation gamma rays from other energies a hig, h resolution Ge(Li) or Ge detector is useful. Nal can be used in some cases when only N is produced. Because many positron emitters V neutro e produceMe b i n H nca y b dirradiation , care mus e takeb t n to avoid interferences. Some possible interfering reactions are-

109 108 Ag(n,2n) Ag t1/ 2.4= 2 2m m

31 3 P(n,2n) °P t1/2 = 2.50 mm

79 78 Br(n,2n) Br t1/2 = 6.50 mm

5l| 53 Fe(n,2n) Fe t]/? = 8.50 mm

63Cu(n,2n) 9.8= m 0m t 62Cu

Q1 QO * Mo(n,2n o M ) t1/ 15.4= 2 m 9m

121Sb(n,2n) 15.9= m ^ 0 m 121 t Sb

Other reactions leading to 0.511 MeV annihilation radiation are tabulate n [2]i d f .interferinI g isotopes have significantly differen r moreo 3 t ) r (factoo half-live 2 f o r e sth tha, N 3 n interferences may be resolved. For example, nitrogen may be determined in the presence of excess phosphorous by letting the sample5 minute? o t ss0 2 decabefor r fo y e countin e 0.51th g V 1Me gammas. Essentiall P will l decaal y y awa n thii y s time, leaving only the 10-mmute N activity. In other cases the interfering isotop y havma ee gamma ray f otheo s r energies wMch may be used to calculate tne contribution of the interfering isotope to the 0.511 MeV peak.

Samples containing hydrogen plus carbon and/or oxygen may exhibit an interference due to the proton recoil reactions C(p,nd an 0(p,a N ). N Contribution) s from these reactions are generall d correctionan w lo y e possiblar s e [2].

Nitrogen detection limits for the usual 0.1 to 10 g sample size e abou ar s1 wt*. 0. t . Analysis times rang0 1 e o frot 5 m sample r hour straightforwarpe s fo r d r applicationpe 2 o t 1 o t s hou morn i ° e complicated cases where interference correctione ar s required.

One of the earlier applications of 14 MeV NAA for dete^mination of nitrogen was determining the protein content of g^am products [29]. The matrix elements (C,H,0) do not int°~fere makin e applicatioth g n very straightforward. This methos i d rapid and precision is 0.3 to 0.5*. Determination of protein by

171 this method has also been reported in other food stuffs [30], plant samples [31,32] and Chinese medicines [33]. Other applications are the determination of nitrogen in Si N. and nitrided glasses [3^], coal conversion liquid fertilized s an [35] r [36].

Competing Techniques for the Determination of Nitrogen Nitrogen may be determined in metals by inert gas fusion and vacuum fusion methods [27] similar to those described earlier for oxygen. LECO makes a combined nitrogen/oxygen analyzer that determines oxyge s describea n d above .s passe ga Afte e sth r through the infrared detector for measurement of oxygen as CO , it is passed through absorber columns to remove HO and CO , then ta thermao l conductivity detecto measuro t r e remaininth e . N g Detection limits are about 0.1 yg/g using sample sizes on the order of 1 gram. This is much more sensitive than the 14 MeV s probablmethoi d e besan d th y t metho r nitrogefo d n i n appropriate metal samples.

The elemental analyzer described above for oxygen in organic alss i s o applicabl r nitrogenfo e . Nitrogen contents a s low as 0.05 to 0.1 wt.î may be determined on samples that typically weigh . abouSommg 3 et inorganic n als ca e analyzedsb o , but care mus e useb t d since many metals interfere wite th h analysis. Several samples per hour can be run. The method has comparable spee d precisioan d d bettean n r sensitivityV Me tha 4 1 n NAA. The method is destructive, and the small mass of sample analyzed could present a problem if the sample is not homogeneous at the 3 mg level.

The classical chemical technique for the determination of nitrogen is the Kjeldahl method [37]. This basically consists of dissolving the sample in sulfuric acid, converting the nitrogen to ammonia. Excess sodium hydroxide is added and the ammonia distille d collectean d a know n i dn amoun f acido t . This solution is back-titrated to determine the amount of ammonia present. The method works for nitrogen compounds such as amines and amino r otherfo t sno acid suct s nitratebu a sh d nitritesan s . Adapta- tion d variationan s e Kjeldahth f o s l method have been developed for some of these latter sample types. One example is reduction of nitrate to ammonia with Devarda's alloy [38]. Many other chemical and instrumental (e.g., ion selective electrode and ion chromatography) methods have been reported in the literature. They generally are applicable to nitrogen in t all,no some t chemicabu , l forms; requirn i e sample b th e o t e solutio d requiran n e several hour r samplpe s o performt e . Chemical methods use standard chemistry lab equipment that is relatively inexpensive. Accuracy and precision are comparable to V NAA1Me ;4 detection limit e lower ar se method Th . e ar s destructive.

Nitroge n somi (aned cases oxygen y als )ma e determine b o d in some sample type y severab s l mass spectroscopy techniques (e.g., glow discharge mass spectroscopy [39], laser mass spectroscopy [40]); ion chromatography [41] and, with some special sampling devices, inductively coupled plasma atomic emission

172 spectroscopy (ICP-AES) [42]. Application of these techniques is still limited. Some require conductive samples while others sample only a micro volume; all require expensive equipment (>US$ 100,000). Nonp strictlu et woulr nitrogese fo ye b d r oxygeo n n determinations.

Silicon Silicon is determined by 14 MeV neutron activation analysis using the reaction:

pu pQ = 2.2 n 5 mi Si(n,pt GammI A ) a ray: 1.778 V (100?8Me ) Escape peaks: 1.267V 8Me 0.756V 8Me Gamma ray counting is best done using a high resolution detector [Ge or Ge(Li)] to resolve the gamma rays of interest from the other gamma rays produced from most matrices. Interfering reactions are:

27Al(n,Y) d an 3128P(n,a)Al 28Al.

The latter reaction has a cross section of 0.5 mb versus 250 mb pQ e reactioth r fo n. witThuSi e aluminu " th sh m interferencs i e onl a probley m when aluminu n greai s i tm exces o silicont s . Other gamma ray lines from other reactions with aluminum [e.g., (n,p) reaction to Mg or (n,a) to *" Na] may be used for interference correction e crosTh . s e phosphorusectioth r fo n s interferenc s comparabli e e reactioth o thar t e fo tn with silicon making this interference very serious. There are other gamma rays from other activation products of phosphorus that can be use e phosphoruo correct dth r fo t s interference. Detection limits for silicon by 14 MeV NAA are on the order of 100 pg, and reported precision e typicall w ar percens fe a ± ty relative.

Applications of 14 MeV NAA to determination of silicon have been most common with geological samples, especiall n casei y s wher e samth ee sample wil e analyzeb l y otheb d r techniquer fo s other elements [11,40,43,44]. Thi s especiallwa s y trur lunafo e r sample d meteoritean s s where sample e limitedar s e otheTh . r advantag f doino el analyse al e gsam th e n o ssampl s thai e t geological samples can be very inhomogeneous, especially for some trace elements that may concentrate in a few grains of a rare phase. Since data interpretation often requires comparing ratios f elementso s importani t i , t thal dat al e trepresentativb a f o e the same sample. 14 MeV NAA has also been used to determine concentrations of Si in iron and steel [45], coal [18], bauxite [46], fly ash and slag [34], mine flue dust [47], wear metal silicon in lubricating oil [48] and biological samples [49]. It has also been used to measure the thickness of silicon films on glass plates [50].

Competing Techniques for the Determination Silicon For many years the main competing techniques for Si determination wert chemicawe e l methods suc s gravimetria h c determina-

173 tion following acid dehydration, colonmetric or gravimetric determination using silico-12-molybdate or titrimetric or gravimetric determination of the (SiF,) ion [2]. All of these are time consuming, destructive and subject to many interferences. Detection limits and precision are better than 14 MeV NAA; accurac comparableis y . Instrumental methods suc atomias h c absorption spectroscopy and ICP-AES are fast, accurate and more sensitive than 14 MeV NAA once the sample is in solution. X-ray fluorescence (XRF s frequentl)i y applie o roct d k samples, though the samples generally are crushed and often fused. Access to proper standard s especialli s y importan r XRFfo t . Electron microprobe is commonly used to determine the silicon content of small volumes of individual mineral grains. 14 MeV NAA would be quicker and easier to perform for bulk silicon determinations than any of the competing methods.

Some Other A ApplicationNA V Me 4 1 f o s Fluorine may be determined by 14 MeV NAA using the reactions :

19F(n,2n)l8F t = 112.0 min or:

F(n,p)G t1/ 2= 27. 1s

In addition the (n,a) reaction on F produces the same N isotope used for the determination of oxygen. In oxygen-free samples ,n excellena thi s i s r tfluorin fo reactio e us eo t n determination. Detection limits of less than 100 microg^ams of fluorine are possible. Since many fluorine compounds are difficult to dissolve and difficult to analyze by other methods, 14 Me\/ NAA can be a very useful tool. For example, the teflon con- Ion-graphitf te a ten f o t e composit1 NA s determine V wa e Me 4 1 y b d [34]. This sample would have been very difficult to dissolve for more conventional analysis. Some other applicationV Me 4 1 f o s NAA for F are: bones [51], barite ores [52], and polymers [53]. X-ray fluorescence is capable of some analyses for fluorine if proper standards are available. Both ion chromatog^aphy and ion selective electrode e use b o determin t dn ca s e fluorine concentrations in some solutions. 14 MeV is perhaps unique in being a nondestructive instrumental method for determination of fluorine.

Other elements with goo A dinclud NA sensitivit V Me e 4 1 y b y Mg, AI, P, Sc, V, Fe, Cu, Zn, Ga, Sb, Ba and Hg. Detection 0 microgram20 o t 0 e rangr 1 eacth fo sf f o hn o elimiti e ar s these. Many other elements have easily detected and measured products when presen n largei t r amounts. Sone other applications from the recent literature include Sb and Cl in synthetic rubbers [54]n coai l,S n cas[55]i tg r bauxit ;M i iro l A n d e[56]an i ,S [46]; F and P in bones [51]; Pb in aquatic plants [57]; Cu, Si, Al, Mg and Mn in iron artifacts [58] and Cu and Ag in ancient coins [59,60].

174 THE FUTURE PLACE OF 14 MEV NEUTRON ACTIVATION ANALYSIS 14 MeV NAA should continue to be a method of choice for determinatio f oxygeo n r nitrogeo n mann i n y samples. This i s especially trur productiofo e r manufacturino n g situations where many w analysetypefe f materiala o s f o s e requiredar s e Th . nondestructive, minimal sample preparation feature e techth -f o s nique make it very rapid, especially compared to methods that require sample dissolution but also compared to instrumental methods that require vacuum, sample homogenization or surface preparation A lendNA v s Me itsel 4 1 . f wel o automatet l d sample handling and data reduction. The one situation where 14 MeV NAA would be a second choice method is determination of oxygen and/or nitrogen in iron, steel and other metals where inert gas fusion is much more sensitive and comparably rapid. Because of the cost of acquiring a 14 MeV neutron generator and setting up a lab to house it and use radioactive materials, most other application V wilMe l4 1 probabl f n o facilitiesi e b y s doing other type f woro s k with radioactivity. These labs already have the administrative procedures for radioactive sample handling in place and have a personnel base already familiar with radiation handling, counting equipment and basic nuclear science. As with most instrumental analytical techniques, the basic principle of the method sounds simple, but application to a new type of sample requires skill and knowledge to avoid or manage possible interference d othean s r experimental pitfalls. Except for the repeated analysis for oxygen, nitrogen, or silicon in a production situation, there is probably no single analytical need to justify setting up a new 14 MeV neutron activation facility. The technique does continue to be a useful in research and educational laboratories wher a large e variet f samplo y e types s especialli A e analyzedNA ar V Me y 4 usefu1 . r har o fo dislt d - solve samples, composites and halogen-containing samples. The generatof bein o e adde e th a neutrog us ds ha rr o n V sourcMe 4 (1 e thermalized) for other nuclear physics or nuclear analytical applications (e.g., cross section determinations, radiation effects, shielding studies, neutron induced particle track production). The generator can also be used in a university settin a majo s a gr teachin e user b isotop fo dgn ca d tool an t eI . tracer production, neutron activation analysis, neutron radiography and other uses, in a similar way to a teaching reactor, at a very small fraction of the cost of even the smallest reactor facility. Severa w developmentne l e expandinar s e applicationth g f o s 14 MeV NAA. The development of portable 14 MeV generators has allowed 14 MeV NAA to move into the field and into the plant. Generators and gamma detectors have been built into down hole probes for well logging and geochemical exploration [61,62], Others have been buil r medicafo t l application r monitorfo d an s - ing flod contenan w n plani t t feed lines [63]. Several special high flux 14 MeV neutron generators capable of producing fluxes t sampla s e irradiatio m c n 0 1 n x position 2 o ut p s have 12 -2 -1 recently been built [64,65]. These permit determination of con- 0 centrationelement3 1 ug/ o t e a samplg th n p i u level st a ef o s . Some new designs for 14 MeV generators are also being produced.

175 These include radial geometries, cyclotron sources and quadrapole generators [66]. These have capabilities for higher fluxes, smaller size and/or simpler operation than the traditional deuteron accelerators described earlier. Successful development of these could open new applications for 14 MeV NAA.

Acknowledgement . Chambersm gratefuB a . I W s: o t . GonzaleslM , , . Procto. M Kelly J . Merril. M P r . sharinM . fo R ,rd an l g their experiences with "competing methods".

REFERENCES

[I] LUTZ, G.J., Ed., 14 MeV Neutron Generators in Activation Analysis A Bibliography: , (Tech. Note 533), U.S. Nat. Bur. Stand., Washington, D.C. (1970). [2] NARGOLWALLA, S.S., PRZYBYLOWICZ, E.P., Activation Analysis with Neutron Generators, John Wiley w YorNe ,k (1973). ] WOOD[3 , D.E., "Activation Analysi sV Neutro Me wit 4 1 h n Generators", Advances in Activation Analysis 2_ (1972) 265. ] EHMANN[4 , W.D., YATES S.W., Anal. Chem 8 (19865 . ) 49R. ] DeSOETE[5 GIJBELS, D. , HOSTE, R. ,Neutro, ,J. n Activation Analysis, John Wiley (1972). [6] WICKER, E.E., "Activation Analysis", Determination of Gaseous Elements Metals (MELNICK, L.M., LEWIS, L.L., HOLT, B.D., Eds), John Wiley, New York (1974) /5. ] McKLVEEN[7 , J.W., Fast Neutron Activation Analysis Elemental Data Base, Ann Arbor Science, Ann Arbor, MI (1931). [8] Handbook on Nuclear Activation Cross-Sections (Teen. Report . 156)SerNo . , IAEA, Vienna (1974). [9] ERDTMANN, G., SOYKA, W., The Gamma Rays of the Radionuclides, Verlag Chemie, New York (1979). [10] ERDTMAN , PETR G. , N"Nuclea H. I r Activation Analysis: Fundamental d Techniques"an s , Treatis n Analyticao e l Chemistry Part I ?nd ed- Volume 14 (ELVING P.J., KRItfAN V., KOLTHOFF I.M., Eds.), John Wile w YorNe yk (1986) 419. [II] WANKE, H. et al., J. Radioanal. Chem. 38 (1977) 363- [12] STEWART, D.C., Handiling radioactivity, John Wiley w YorNe , k (1981). [13] NASS H.W. . InorgJ , . Nucl. Chem 3 (193 . 71) 617. [14] NAGASHIMA T. et al., Yogyo Kyokaishi 88 9 (1980) 511. [15] BILD R.W., IEEE Trans. Nucl. Sei. NS-2 2 8(1931 ) 1622. [16] VOLBARTH, A. et al., Chem. Geol. J_9 (197D 327. [17] VOLBARTH, A. et al., Fuel 5J7 (1978) 49. [18] HAMRIN, t al.C.Ee . ,JR . FUE 8 (19795 L . )48 [19] MAHAJAN, OM. P., Fuel 6_4 7 (1985) 973- [20] KHALIL, S.R. et al., J. Radioanal. Chem. 57 (1980) 195. [21] CHIBA , ANDO , M. Nippo, T. , n Kmzoku Gakkaish 7 (1982 6 _4 i ) 687. [22] ANDO , CHIBAT. , , BunsekM. , i Kagak 5 (1982 _ 3j u ) 294. [23] CHIBA, M., ANDO, T., Bunseki Kagaku ^]_ 4 (1,9S2) 179. [24] KIREIKO V.V t al.e . . LesJ , s Comm. Met (19867 11 . ) 24~. [25] SZEGED , DIVOS. RadioanaJ S. I , F. , l Chem2 (198^ _ 8j . ; 317. [26] BRQVKO, A.P .t al.e , Zavod. Lab2 (1985 1 5 .- )53

176 [27] MARTIN, J.F., MELNICK, L.M., "Vacuum and Inert-Gas Fusion", Determinatio f Gaseouo n s Elements Metals (MELNICK, L.M., LEWIS, L.L., HOLT, B.D., Eds), John Wiley w YorNe ,k (1974) 113. [28] SILVERMAN, L., LaMONT, B.D., "Oxygen", Determination of Gaseous Elements Metals (MELNICK, L.M., LEWIS, L.L., HOLT, B.D., Eds), John Wiley w YorNe , k (1974) 341. [29] WOOD, D.E., "Fast Neutron Activation Analysi r Nitrogefo s n in Grain Products", KN-65-186 Kaman Nuclear Corp., Colorado Springs, CO, (1965). [30] SAMEI, M.B. et al., J. Radioanal. Nucl. Chem. 103 2 (1986) 81. [31] SOTO MORAN, R.L., SZEGEDI Radioanal. J , ,S. . Nucl. Chem6 9 . 2 (1985) 69. [32] NDIOKWERE JERABEK, L. . Talant, CH , P. ,5 (1983 0 3 a ) 377. [33] CHUANG, L.S., CHIU, W.Y.. ChinJ . ., Am (19802 Med1- 8 .) 47. [34] BILD, R.W., unpublished data. [35] EHMANN, W.D. et al., J. Radioanal. Chem. 59 2 (1980)407. [36] KAFALA, S. et al., J. Radioanal. Nucl. Chem. 92 2 (1985) 341. [37] BEEGHLY, H.F.,"Ni trogen", Determination of Gaseous Elements Metals (MELNICK, L.M., LEWIS, L.L., HOLT, B.D., Eds), John Wiley, New York (1974) 321. [38] HILLEBRAND, W.F., LUNDELL, G.E.F., BRIGHT, H.A., HUFFMAN, , ApplieJ. d Inorganic Analysis, John Wiley w YorNe , k (1953) 786. [39 G INDUSTRIES]V G 900"V ,0 Applications Update IsotopeG V " s Ltd., Cheshire, England, (1986). [40] SHANKAI, Z., et al., Anal. Chem. 56 3 (1984) 382. [41] MERRILL, R.M., "Ion Chromatography", Metals Handbook Ninth Edition Materials Characterization Vol. 10, American Society for Metals, Metals Park, OH (1986) 658. [42] FARIES, L.M., "Inductively Coupled Plasma Atomic Emission Spectroscopy", Metals Handbook Ninth Edition Vol. 10 Materials Characterization, American Society for Metals, Metals Park H (1986O , . )31 [43] MORGAN, J.W., EHMANN W.D., Anal. Chim. Acta _4_9 (1970) 287. [44] LAUL, J.C., Atomic Energy Rev3 (1979 7 j_ . ) 603. [45] VANGRIEKEN, R. et al., J. Radioanal. Chem. 6 (1970) 385. [46] LIN, C.-Y., et al., Radiochem. Radioanal. Lett. 53 4 (1982) 203. [47] KAFKA, D., et al., J. Radioanal. Chem. 57 1 (1980) 211. [48] D'AGOSTINO, M.D., KUEHNE, F.J., "The Use of a Simplified Neutron Activation Technique for Analyzing Metallic Wear from Aircraft Hydraulic Systems", Radioisotopes for Aerospace (DEMPSEY, J.C., POLISHUK, P., Eds.), Plenum Press, New York (1966) 346. [49] SCHRAMEL, P., J. Radioanal. Chem. 3 (1969) 29. [50] JAYANTHAKUMAR, S.S., BHORASKAR, V.N. . RadioanalJ , . Nucl. Chem. 1 (1986Lett 4 10 . 1 ) [51] CHINDHADE, V.K. et al., Radiochem. Radioanal. Lett. 52 3 (1982) 141. [52] WANG , HejishQ. , (19843 u^ . )46 [53] JOGLEKAR, A.G. et al., Radiochem. Radioanal. Lett. _45 3 (1980) 199. [54] BILD, R.W. . RadioanalJ , .(19822 Chem1- . 2 )23 7 . [55] KLIE J.H., SHARMA H.D., J. Radioanal. Chem. 71 1-2 (1932)299.

177 [56] KWIECINSKI.S. . RadioanalJ , .(19822 Chem1- . . )17 72 . [57] CHINDHADE, V.K., et al., Radiochem. Radioanal. Lett. _49 5 (1981) 287. [58] BHORASKAR, V.N., et al., J. Radioanal. Nucl. Chem. Lett. 95 2 (1985) 73. [59] ELAYI, A.C., et al., J. Radioanal. Nucl. Chem. 90 1 (1985) 113. [60] CHALOUHI al.t e . ,CH , Nucl. Instr. Meth 0 (198220 . ) 553. [61] HALL, J.E. et al., IEEE Trans. Nucl. Sei. NA-28 2 (1981) 1680. [62] JENSEN, D.H. et al., IEEE Trans. Nucl. Sei. NS-28 2 (1981) 1685. [63] ROCHAU, G.E. et al., IEEE Trans Nucl. Sei. NS-28 2 (1981) 1658. [64] WILLIAMS, R.E t al.e . Radioanal. J , .(19822 Chem1- 2 )7 . 187. [65] PEPELNIK t al.e . ,. RadioanalR ,J .(19822 Chem1- 2 )7 . 393- [66] STOKES, J.A. et al., IEEE Trans. Nucl. Sei. NS-30 (1983) 1623.

DISCUSSION

. HoffmanE e portabl. e fluth d costh Whaan f xo s t i te neutron generator?

R.W. Bild. Portable generators I know of have been either prototype designs r welo l logging tools don'I . t kno realistiwa c cosr eithefo t r type.

Neutron outputs are very dependent on design. Some pulse generators could put 14 t aboutou 0 1 neutron r puls mighpe t s bu e t hav a lifetime f leso e s tha0 1 n pulses. Other pulse generators produce orders of magnitude fewer neutrons but have lifetime f thousando s f pulseso s . Steady state generators coule b d 8 10 made 0 1 neutron.o c coul t se Output r d0 1 pe sprobabl f o s e b y achieved. A practical portable generator for activation analysis has yet to e builtb .

R. Rosenberg. What kin f detectioo d n limi r fluorinfo t uraniun i e m woulu yo d predict?

R.W. Bild. I would expect the detection limit to be 100 to 200 micrograms. The uranium matrix should not be a particular problem in spite of the fissions that will be induced.

178 NUCLEAR ANALYTICAL METHODT SA ELECTRICAT NT L COMMUNICATIONS LABORATORIES Substoichiometric Analysis and its Applications

T. SHIGEMATSU Electrical Communications Laboratories, Nippon Telegrap Telephond han e Corporation, Tokai-mura, Ibaraki-ken, Japan

Abstract

This paper describe e nucleath s r analytical methods thae ar t used at NTT Electrical Communications Laboratories. In particular, e activatioth f o e nius t methoe focuseth n n substoichiometni o ds c s analysiapplicationsit d an s e accuracTh . d precisioan y f o subn - stoichiometnc analysi e investigatear s y considerinb d e effectth g s on systemati d accidentaan c l errors cause n analyticai d l procedures. It is shown that the total errors of substoichiometnc analysis are dominate y statisticab d l fluctuation n activiti s y measurementss i t I . also shown that substoichiometnc analysis is one of the most accurate analytical methods. Thie s characterizatioi th usefus n i l n of electrical materials and devices and in the standardization of other analvtica! methods.

1. Introduction Accurate and precise anal>tical methods are necessar\ in order o t characteriz d develoan e p variou w electricane s l materiald an s devices. Because of this, different types of nuclear analytical methods are being developed at NTT. Nuclear analytical methods are suitable for determining major and trace elements. Trace determination f o lighd sheav an t > elements have been confirme\ b d charged particle and neutron activation analysis. Major and minor elements have also been determine y radiometrib d d isotopan c e dilution analysis. Substoichiometri e mosth f tc o accurat analysie on s i se nuclear analytical methods because e contentb determine n ca s y usinb d g only radioactivity measurements that use simple ehe.meal separations. This can also be done without calibrating chemical yields. Thus, substoichiometnc analysis is a very useful method in obtaining relations between the chemical compositions and the properties of materials. In addition, Substoichiometric analysis is also use n ordei d o standardizt r e other analytical methods. This paper will describe various nuclear analytical methods. In particular it will focus on the accuracy of Substoichiometric analysis. It will be shown that this is one of the moat accurate analytical methods.

179 2. Nuclear Analytical Methods at NTT ECLS Activation analysis, radiochemicad an ) l A analysiRC ( s isotope dilution analysis ( IDA ) have been used for the determinatio f heavo nd lighan y t element n semiconductorsi s , superconductors and optical fibers[l,2]. A summary of these analyses is shown in Figur . 1 Trace e amount f heavo s y elements have been determined by neutron activation analysis usin e nucleath g r e reactorth n i s Japan Atomic Energy Research Institute. A small cyclotron has been used since 1984 for the trace determination of light elements by charged particle activation analysis[3]. RCA and IDA have been applied to the determination of major and minor elements. Radioisotopes have been produce y neutrob d n irradiatiod an n charged particle bombardment.

Element Machine Analytical method

Heavy Nuclear NAA —— -, \ nstrumental element reactor / analysis

(RI) RCA y IRI» 1 DA --~4y , Substoichiometric analysis Light Cyclotron CPA element

RBS

Fig.1 Nuclear analytical methodT NT t a s

In order to achieve simultaneous multi-element determination, instrumental analysis has been developed in both activation analyses. Substoichiometr s beeha ny introduce l thesal o et d analysed an s Substoichiometric analysi s beeha s n applie o about d 3 elements3 t .

. 3 Principle f Substoichiometrio s c Analysis Many type f o Substoichiometris c analysis have already been developed. These include the direct method, the RI addition and the activation methods. In the direct method, a large excess and known f o carrie ) amoun e s activatei addeM th r ( o tt d d sample. A known amount ( m ) of element is separated substoichiometn- s i s it measured) activit d x calla an e ( activityys Th i . x A y : I obtaine . Eq y b d

Ax = ( M / m )ax ( 1 ) e direcTh t metho e determinatio th s i usefu dr fo l f o inducen d activity. e I activatioth additioR d e th an In n methods a know, n amoun f standaro t d sample shoul e preparedb d . d Tesstandaran t d samples ( Contents; MX, Ms ) are labeled by adding carrier free

180 I R additio e th R( n Activityn Ii method ) A , . Definite amounts ( m ) are separated substoichiometrically from both samples and e theimeasuredar ) r activities e a contens i Th calcu, . ax X M ( s-t late y Eq.2b d : MX = ( as / ax )Ms ( 2 )

These principles are shown in Figure 2.

a- A, (S.3)- i

M =— x m Direct method

Mx / M« Carrier free RI(A) A '1 I

Ms M5 "MS RI addition method =^ M

i 11 r-. vmAx l ~ 1T~ _ im i -(s.s)- 1 / m._ J X M M| X m M Activation Carrier:M M»M ,M X s as T r^ mh \ 1 V/A | 1_ (s.s)-* 1 Ms Ms W, J m M

Activation method

Fig.2 Substoichiometiric analysis

e principl e Th activatioth f o e n e consideremethob n ca d n i d more detail. Tesd standaran t d samplee ar ) ( Contentss Ms , MX ; activated simultaneously unde e samth r e conditions. Thene th , specifi c) equal e activitX testh M s t f / o thay samplf x o A t ( e the standard sample ( As / Ms ). Large excess amounts ( M ) of carrie e addear r o t botd h samples, followee pre-separatioth y b d n of objective elemen n ordei t o t removr e interferinth e g activities d elementsan e carrieTh . r amoun ( ) Activitiesvarie s M A t s , Ax ; by M' ( Activity; Ax1 ) in the test sample and by M" ( Activity; As' ) in the standard sample. Content MX is usually calculated by Eq.3:

) ( Ax'M M3 = X ( / " As'M s M ) 1

M' and M" should be determined in this case.

181 In substoichiometnc analysis, small and definite amounts ( m ) are separated from M' and M" and their activities ( ax, as ) e ar measured. e determineThenb n ,ca y contenEq.4 b X d M : t

MX = ( ax / as )Ms ( 4 )

s i cleaIt r from these equations that e detercontenb n -ca t mined by using only activity measurements. There is no needs for correction f o chemicas l yield n i substoichiometns c analysise Th . principl e activatioth f o e n metho s showi d n Figuri n . 3 e

Test sample

Standard sample

Fig.3 Errors in substoichiometric analysis (Activation method)

. 4 Error n Substoichiometni s c Analysis Three kinds of systematic errors and an accidental error are considered in the chemical procedures of substoichiometric anahsis. These errors should be decreased to establish more accurate analytical methods. s § contaminatd an x S e d standartesIth an f et d samples before substoichiometric separation, specific activiu ( Ax / M ) of the test sample changes to ( Ax / ( M + S \ ) ), and that of the standard sample ( As / M ) changes to ( As / ( M + S s ) ). This systematic error is caused by variations of specific activity with contamination. There are two other kinds of systematic errors that affect e n i th substoichiometriquantit) m ( y c n separationa s i e On . error (4m ) caused by co-existing elements. Their formation constants e substoichiometriwitth h c reagen e largear t r than those e objectivth of e element e otheTh . r (/I s i causeS ) y reactivitb d y r o instabilit (e reagentth f o ) y . An accidental error is caused b> statistical fluctuations in activity measurements.

. 5 Accurac f Substoichiometno y c Analysis The systematic errors caused by variations in specific activities and by interference of co-existing elements in the activation e methotreatee b sam th n n eca i d \\ay. Systematic error cause y variationb d n specifii s c activity depende quanth n -o s tities of contamination ( S s, § x ) and their ratio (5s / §x ). s / {5"x is usually less than 1.0. The quantities of contamination can e decreaseb d belo 1 jjgw . 0.1- f carrieo g s m i 1usuallr y added e sampleth o t , therefore e erroth , r will remain below 0.01e .Th error caused by interference of co-existing elements decreases becaus f pre-separatioo e d becomean n s negligible. Thi s i shows n i n Figur. 4 e

0.01 -

-0.01-

0.000 0.005 0.010

Fig.4 Errors in subtoichiometric analysis

These results indicate that it is necessary to add as large a carrie s a ro possibl t trea d t an ebot h d standartesan t d samples under the same conditions in order to establish more accurate activation method. Reactivit r o instabilite ( substoichiometrith y f o ) y c reagent causes systematic error s a shows n n i thiFigurI n s. 5 casee th e

0.01 -

0.001 _6 -5 -4 -3 log (Bi/g) Fig.5 Erro n reactiviti r f o y dithizone with bismuth 3 0. 0.8,2= = 1 0.5X X ) = ) X )3 ,

183 substoichiometric extraction of bismuth with dithizone is considered e abscissTh .s i bismuta he amounth weight f I f o t . bismuth is more than 10 jag, the error caused by reactivity of the reagent decreases below 0.1 %. This is because a reagent wit a hlarg e extraction constan e substoichiometrith s i user t fo d c extraction. This typ f erroo e s i alsr o negligible. The effect of accidental error in activity measurements are show n i Figurn . 6 Whe ee totath n l activit^ f o Disnuty0 1 s i h counts and the extraction ratio ( x ) of bismuth is 0.5, accidental error reaches 1.4%. These results indicate s easi thao decreast t yi t e systematic error e activite sseparateTh th . f beloo % y 1 wd part shoule b d more ^ thacount10 n n ordei s o decreast r e total errors belo. 1% w s i confirmeIt d from these results that substoichiometric e mosth f t o accurat analysie on s i se methods.

0.001 0.0

Fig.6 Error in activity measurement

a) 10*counts 1) Bi b) 105counts 2) 3}

. 6 Criterio f Substoichiometrio n c Analysis If substoichiometric separation for an element is established, the RI addition and the activation methods can be applied according to the content of the element. The activation method is used for trace determination and the RI addition method is used for determination of more than 10 ^g of element. The sample weight also depends on the analytical method. This is shown in e criterioTh Figur f . o 7 substoichiometrine c analysi s i shows n i n Figur . 8 e I Generall}additioR e nth , metho s i usefudr fo l determination of major and minor components. The activation method is used for trace determination. However, the RI addition metho s i used n determinini d g trace element r largfo s e amountf o s test samples e activatioTh . n metho s i adequatd n i determinine g majod minoan r r element r smalfo sl test samples. Elements that can be separated substoichiometncallv are shown in Figure 9. The\ mak t i possiblee substoichiometrith o t e c anal>si 3 elements3 f o s .

184 Activation RI addition

Activation -

Carrier IV!

Substoichiometric

(MX«M) separation

Fig.7 Substoichiometric analysis

0.0001 0.001 0.01 0.1 1 Sample weght,(g) Fig.8 Criterion of analytical method

1 U III IV V VI VU VIII

1 H He

La Séries (La) Ce Pr Kd) Pm Gd (Tb) Dy Ho Er Tm Lu @)

Fig.9 Substoichiometric separatio f elementno s

185 . Substoichiometn7 c Analysi d othean s r Analytical Methods Substoichiometnc analysis is used for calibration of other non-nuclear analytical methods. Thi s i illustrates e casth f o en i d synchrotron radiation x-ray fluorescence analysis ( SR-XRF ) for determination of major components of a superconducting oxide BaPb,- . a ) Thithi s i nB sx BiBP filx O( m a crystal grow n SrTiOo n . Reference samples were preparee calibratioth r fo d f SR-XRFo n . The compositions of the reference samples were determined by the RI addition method within errors of 1%. In particular, this was done for lead and bismuth.

8. Conclusion It is important to develop more accurate nuclear analytical methods. It was shown here that Substoichiometnc analysis is one e mosth of t accurate analytical methods. Thi s i importans o t t characterize and develop new electrical materials and devices. This can also be applied to the standardization of other analytical methods.

REFERENCES

[1] KUDO , SUZUKIK. , , TrendN. , s Anal. Chem., 3(1984)20. [2] KUDO, K., SUZUKI, N., J. Radioanal. Nucl. Chem., 88(1985)75. [3] SHIKANO, K., YONEZAWA, H., SHIGEMATSU, T., J. Radioanal. Nucl. Chem., Lett. 95(1985)55.

186 NUCLEAR ANALYTICAL METHODS IN STANDARDS CERTIFICATION

R.M. LINDSTROM Cente r Analyticafo r l Chemistry, National Bureau of Standards, Gaithersburg, Maryland, United State f Americso a

Abstract Nuclear method f chemicao s l analysis have playe a prominend e taccurat th rol n i e e determination of the composition of Standard Reference Materials and other standards. The characteristic sensitivity, multielement capability, and freedom from chemical biases unique to nuclear methods assure them a continuing role in the standards laboratory.

f nucleao e us r e analyticaITh . l method r standardfo s s analysis One of the few products (as distingished from services) of the U. S. National Bureau of Standards is a variety of Standard Reference Materials (SRMS), certifie e Bureath y b ud o possest s certain physical propertie a specifie r o s d chemical composition. Nuclear methods of analysis play a prominent role e certificatioith n n process [1,2]. Unles n unquestionabla s y reliable "definitive" method of analysis can be used, concentrations in SRMs are certified by a combination of two or more analytical methods which are as different as possible e sourceith n f theio s r characteristic errors. Nuclear methods, largely instrumental thermal neutron activation analysis with gamma-ray spectroscopy but also including some others, are of particular value in this work because of some of their distinguishing characteristics: 1. Substantial freedom from systematic errors e physicaTh . l processes involved in nuclear methods are well understood. Radioactive growth and decay are rigorously exponential. The number of energetically possible nuclear reactions fro a givem n target nuclid r moderat s smallfo i e d an ,e irradiation energ l possiblal y e reactione b n ca s enumerated by inspection of a table of nuclides. High-resolution gamma-ray spectroscopy affords qualitative identificatio e nuclideth f o n s presen s wela ts theia l r quantitation. The density of known gamma-ray lines in energy space is small compared with the resolution of modern detectors t manno yd analyticallan , y important decay gamma rays remaie discoveredb o t n e presencTh . f o e possible interferences may often be readily tested when multiple lines of either component are emitted.

187 . 2 Complementarit o othet y r methods .A differen t suitf o e elements is measurable by using nuclear rather than chemical reactions e detectioth d an , n limite ar s quantitatively different. Equally important, the kinds of error o whict s h nuclear methodo t e e subjecdu ar s e ar t different physical phenomena and are therefore likely to giv a differene e tresults th bia n i s . 3. Freedom from analytical blank and other problems related o dissolutiont . Excep r pre-ifo t r radiation handlind an g packaging, o therreagenn s i e te usua blanth n li k sense. Analytical methods wETch require thae analytn th ti e b e solution call for chemical laboratory skills far beyond simple weighin d packagingan g . With some matrices (rocks containing complex silicates like chromit r zircono e ; fatty animal t tissues)eas no o assur t ys i t ei , complete dissolution without using extreme conditions or multistep chemistry. 4. Quantitatively known precision. The random process of radioactive decay givea prior n a s i estimate th f o e variation to be expected between samples. A simple t test [3] shows immediately whether counting statistics is the limiting factor in precision. The accuracy can be comparabl e precisioth o t e t levela n s well belo% 1 w [4,5,6], even for a decaying source. 5. Multielement capability. Gamma-ray spectroscopy is inherentl a multinuclidy e analytical processe th , components of which add linearly. Radioactive decay adds the dimension of time, which can often act as a perfect separations chemis o resolvt t e otherwise interfering components such as Cr-51 and Ti-51. 6. Sensitivity. Neutron activation analysis in particular has been shown to be applicable to the analysis of many element t sub-picograa s m amounts (7,8) e optioTh . f o n chemical separation after irradiation is often available e blank-frefoth r e remova f interferino l g radioactivities. Macroscopic quantities of the sought-for and interfering element e freelb y ma sy adde n ordei d o east e r th e chemistr o facilitatt d an y e measurementh e f chemicao t l yield.

These characteristic f nucleao s r methods have been widely exploited, particularly in research into trace element analytical methodology. Fifty-six percent of all published analyse S multielemenNB f o s t SRMs have been performey b d nuclear techniques, accordin a recen o t g t survey [9], Contemporary trace element geochemistry, from lunar sampld an e meteorite analysis [10] to mineral exploration, relies heavily on neutron activation analysis. Because of its freedom from blank, neutron activatio e mosth ts i npowerfu l technique available for the study of contamination in handling and sampling animal tissues [11].

188 II. Nuclear methods other than conventional activation analysis Nuclear method f analysio s e oldear s r thae 50-yeath n r history of neutron activation, the first recorded tracer experiment following closely the discovery of radioactivity. Since then, natura d artificiaan l l radiotracers have been used in an enormous variety of applications, including engineering studie f metallio s c wear, distributio f soluteo n s between soli d moltean d n silicon e movementh , f oceao t n currentd an s ground waters, and in countless studies of biochemical pathways [12]. Radiotracers can be a great aid in developing procedures to be applied with other analytical methods [13]. e analyticaTh f prompo e us tl nuclear radiatios ha n expanded in recent years. Beams of neutrons from reactors have bee e simultaneounth user fo d s determinatio f twento n r moro y e elements by measurement of prompt gamma rays [14]. Pure beams of low-energy "cold" neutrons, now becoming available at several reactors, offer attractive possibilities for chemical analysis [15]. Rutherford backscattering and the newer technique of neutron depth profiling [16] are used to study the concentratio f elemento n a functio s a s f depto n h near surfaces.

III. SRMs as multielement standards Althoug e oligomonitoth h r metho f activatioo d n analysis a firi sn o mbein t theoreticapu g d experimentaan l l basis, most analyses of high accuracy still employ a co-irradiated standard for each element to be determined. The use of standards has been recently discussed in two publications from this laboratory [1,17]. SRM d Certifiesan froS NB m d Reference Materials (CRMs) issued by other bodies are frequently used as multielement comparator standards [17] n additioi , o theit n r recommended s knowa e nus material o verift s e correctnesr o th y w ne a f o s routine chemical procedure n somI . e ways CRMe suitablar s r fo e this use: thee convenientlar y y availabl n mani e y laboratories, and the analyst's confidence in their compositio e reputatios backei nth y b da majo f o nr institution. Reference materials, are not ideally suited for use as comparators, however. These materials are expensive, and are never produced in a large enough supply to last for many years. Futhermore, CRMs made from geological materials or from plant or animal tissue invariably are made up of more than one solid phase and are thus necessarily heterogeneous at some level [18,19]. Ingamell d Switzer[19an s ] giv n extrema e e example: the expectation value of the number of grains of chromite, which carries at least half the chromium, in a 100-milligram analytical sample of the USGS standard granite G-l is 0.5. Finally, the certified concentrations of at least some components of a multielement CRM are not as accurately e componentknowth s a na syntheti f o s c standard solution that can be made in the analyst's own laboratory. With conventional analytical apparatus, weighing and pipetting can be accurate to 0.2% or better; by contrast, the average stated uncertainty

189 of seven major and minor elements in NBS SRM 1633a Fly Ash is f seventeeo 3.4% d an , n trace elements 4.4%. Only fouf o r twenty-on e certifier bettero ar e % 1 o .t dHowever M whicCR a ,h is homogeneous may be accurately calibrated against primary standards by the user and used as a working standard thereafter [20]. A number of workers have addressed the need for a working multielement calibrant material which is homogeneous, similar in matrix composition to the samples to be analyzed, and of well-known composition. If a liquid or colloid is uniformly doped and then solidified into storable form, a standard can e prepareb d whose homogeneit s guaranteei y d e modmerelth e y b y f preparationo . Many laboratories have used fine silica slurried with solution standards. Photographic grade gelatin [21], urea-formaldehyde polymer [22], and polyacrylamide gel [23] have been used as analogs of plant and animal tissue, while clay [24] and mineral glasses [25,26] have been made as geological materials. Synthetic multielement standards are now under active development in several laboratories.

IV. The future of nuclear methods The well-publicized steep growth in the capability of computers and other electronic devices, along with steeply falling hardware costs, have e economiworketh o t d c advantage of nuclear methods, with the result that the instrumentation for activation analysis is comparable in cost to the newer atomic methods A .4096-channe l pulse height analyzer boughy b t our laboratory in 1974 cost $20,000; for this number of 1986 n entira y bu e n analyticadollarca e on s l system consistinf o g n 8192-channea l analyzer a substantia, l microcomputer0 3 , Mbyte f disko s d softwaran , r datfo ea acquisitio d spectraan n l data analysis. The modern system is more linear, more reliable, more versatile d performan , s much more work thas it n ancestor. Detector e alsar s o more efficient f higheo , r resolution, more portable, and less fragile for a given number of currency units than was the case only a few years ago. Largely becaus f improveo e d instrumentation e numbeth ,d an r quality of analytical measurements per analyst-week in our laborator s increaseha y d noticeabl e pasn yearsth te tn i y . Becaus e simplth e e unit processe f instrumentao s l neutron activation analysis are so easily automated, automatic sample changers have been familia n radiochemicai r l laboratorier fo s decades. The multiplication of analyst and instrument time by the use of robots is only beginning to appear in more classica d complicatean l d chemical applications. Computer softwar r gamma-rafo e y spectroscopye th r fo d an , additional calculations involved in activation analysis, is becoming somewhat standardized since therw e onlfe ar ea y commercial suppliers of specialized computer-based systems for data acquisition. The wide distribution of some of this software, among critical users with different requirements, has led to the elimination of weaknesses in the early algorithms. Some well-documented software created in users' laboratories (Sampo,[27] Hypermet [28]) can now be obtained from equipment manufacturer s wela ss directla l y froe th m authors. This quasi-standardization results in good

190 comparability of data obtained in different laboratories, and make t likeli s y that researchers will find familiar equipment and procedures when working as a visitor in another institution. The most important limitation in the wide application of nuclear methods of chemical analysis in the near future and in the long run may be access to reactors and other sources of nuclear-active particles. A recent compilation listed 61 researc d traininan h d abovan n gi W ek reactor 0 10 f o s operation i n the United States alone [29], mos f whico te ar h suitablr fo e activation analysis and for producing radiotracers, Their number is not likely to increase soon, however. The decreasing necessity for new nuclear electric power plants implies that fewer nuclear engineer e beinar s g traine n feweo d r university reactors. Regulatory changen i s response to concerns of public safety make operation of reactors more and more expensive; the additional capital expenses resulting from the recent requirement that research reactors conver o low-enrichet t d fuey welma l l shut down some marginal facilities. Despite these uncertainties e immediatth , e futurf o e nuclear methods is secure: their special advantages are vital in special situations suc s standarda h s certificationd an , their routine use for many applications such as geochemical d biomédicaan l trace analysis [30 s assured]i n manI . y ways, nuclear methods have set and will continue to set the standard for high-quality elemental analysis, especiall t traca y e levels. Meinke's comment of twenty years ago is still valid: "No quarter wil e giveb l o radiochemicat n lo n method d an - s quarter shoul e askedb d " [31].

References

11] GREENBERG,R.R., presented at MTAA-7, Copenhagen 1986, J. Radioanal. Nucl. Chem. (in press). [2] GILLS,T.E., "Nuclear Methods — An Integral Part of the NBS Certification Program",in Proc. 5th Internat. Conf. Nucl. Methods in Envir. Res. (CONF-840408), (VOGT,J.R., Ed), U. S. Dept f Energyo . , Washington (1984), 634.

[3] HEYDORN,K., NORGARD,K., J. Radioanal. Chem. 1J5 (1973) 683. ] LINDSTROM,R.M.[4 , HARRI SON,S.H., HARRI S,J.M. Appl. J , . Phys. 49 (1978) 5903. [5] FLEMING,R.F., LINDSTROM,R.M., presented at MTAA-7, Copenhagen 1986, J. Radioanal. Nucl. Chem. (in press). ] GREENBERG,R.R.[6 , CARPENTER,B.S., "High Accuracy/High Precision Determination of U-235 in Nondestructive Assay Standards by Gamma-Ray Spectrometry", in Proc. 5th Internat. Conf. Nucl. Methods in Envir. Res. (CONF-840408), (VOGT,J.R.f . DeptS Ed) f Energy. o .,U , Washington (1984), 644.

] ZEISLER,R.[7 , GREENBERG . RadioanalJ , . R . ,R . Chem (1982j TJ . . )27

191 ] LINDSTROM,R.M.[8 , "Activation Analysi f Semiconductoo s r Materials",in Microelectronics Processing, (CASPER,L.A. , Ed), American Chemical Society, Washington C (1986)D , , 294.

(9] GLADNEY,E.S. , BURNS,C.E., PERRIN,D.R., ROELANDTS,I., GILLS,T. E., 1982 Compilation of Elemental Concentration Data for NBS Biological, Geological d Environmentaan , l Standard Reference Materials (NBS Spec. Pub. 260-88), U. S. Government Printing Office, Washington, D.C. (1984). [10] LIPSCHUTZ,M.E .Anal, . Chem , 96958 .A (1986). [11] VERSIECK,J.M.J., SPEECKE,A.B.H., "Contaminations Induced by Collectio f Liveo n r Biopsie d Humaan s n Blood",in Nuclear Activation Techniques in the Life Sciences, IAEA, Vienna (1972). 39 , [12] BOWEN,H.J.M., "Introduction to Radiotracer Experiments" , in Treatise on Analytical Chemistry, Part I: Theory and Practice d Ed.2n , , (KOLTHOFF,I.M., ELVING,P.J., KRIVAN,V., Eds), Wiley w YorNe , k (1986), 193. [13] KRIVAN,V., "Application of Radiotracers to Methodological Studies in Trace Element Analysis",in Treatise on Analytical Chemistry, Part I: Theory and Practice, 2nd ed., (KOLTHOFF,!. M., ELVING,P.J., KRIVAN,V., Eds), Wiley, New York (1986), 339. [14] LINDSTROM,R.M., ANDERSON,D.L., "Analytical Neutron-Capture Gamma-Ray Spectroscopy: Statu d Prospects",ian s n Capture Gamma-Ray Spectroscopy and Related Topics - 1984, (RAMAN,S., Ed), Am. Inst. Physics, New York (1985), 810.

[15] LINDSTROM,R.M., ZEISLER,R., ROSSBACH,M., presented at MTAA-7, Copenhagen 1986 . RadioanalJ , . Nucl. Chem. n pressi , .

[16] DOWNING,R.G., FLEMING,R.F., LANGLAND,J.K ., VINCENT,D . H . , Nucl. Inst. Meth. Phys. Res. 218 (1983) 47. [17] BECKER.D.A., presente MTAA-7t a d , Copenhagen 1986. J , Radioanal. Nucl. Chem n press)(i . . [18] DAMSGAARD,E., HEYDORN,K., REITZ,B., "Determination of Vanadium in Biological Materials by Neutron Activation Analysis",in Nuclear Activation Technique e Lifth en i s Sciences (Proc. Symp. Bled, Apr. 1972), IAEA, Vienna (1972), 119. [19] INGAMELLS,C.O .SWITZER,P., , Talant (19730 2_ a ) 547. [20] KOROTEV,R.L., presented at Halifax Conf., 1985, J. Radioanal. Nucl. Chem. n pressi , . [21] ANDERSON MURPHY,J.J., . H . ,D , WHITE,W.W., Anal. Chem (19768 4_ . ) 116.

[22] MOSULISHVILI,L.M., DUNDUA,V.Y., KHARABADZE,N.E., EFREMOVA,E. Y., CHIKHLADZE,N.V., J. Radioanal. Nucl. Chem. 83 (1984) 13. [23] MATSUMOTO,K., SUZUKI,N., Radiochem. Radioanal. Letters 42 (1980) 99.

192 [24] PERLMAN,!., ASARO,F., "Pottery Analysi y Neutrob s n Activation",in Science and Archaeology, (BRILL,R.H., Ed), MIT Press, Cambridge (1971), 182. [25] MARINENKO,R.B., Preparation and Characterization of K-411 d K-41an 2 Mineral Glasse r Microanalysisfo s 0 (NB47 M S SR : Spec. Pub. 260-74), National Burea f Standardso u , Washington, DC (1982). [26] LINDSTROM,D.J., KOROTEV,R.L., private communication, 1985.

[27] ROUTTI,J.T., PRUSSIN, S. G. , Nucl . Inst. Methods 72. (1969) 125.

[28] PHILLIPS,G.W., MARLOW,K.W., Nucl. Inst. Method 7 (197613 s ) 525. [29] BURN,R.R., BILOF,R.S., Research, Training, Test, and Production Reactor Directory, American Nuclear Society, LaGrange Park, 111. (1983).

[30] DEBRUIN,M., KORTHOVEN,P.J.M., BODE,P. . J Radioanal, . Chem. 7_0 (1982) 497. [31] MEINKE,W.W., "The Plac f Radiochemicao e l Method f Analysiso s : Today and Tomorrow",in Proc. Symp. on Radiochemical Methods f Analysiso , IAEA, Vienna (1965). 13 ,

DISCUSSION

G. s Revelnecessari t I . o remait y n very e prudencertificatioth r fo t f o n very low concentrations. I know several cases where different methods preforme n differeni d t laboratorie e samsth giv er samplesfo e e samth , e e resultresultth d e an determinatios s e th cas wer t th f true o en no eI . f o n oxyge n aluminiumi n , during five year e concentrationth s s considered like true varied from some p.p.m o somt . e t p.p.beaca d h an .step , different methods give the same result at the same time.

My second comment concerns the blank value due to surface pollution for solid samples. Theoretically this blank is avoided in activation methods e cleaninb th e ysurfac th f o g e after irradiation n fac I n ultra-traci . t e analyses e supposeo efficienb ,s n thit ca no s do s easd s a cleanint t an y no s i g (irregular attack, redeposition, phenomena, etc.)

193 ANALYSIS OF DIFFICULT MATERIALS: COMPARISON NUCLEAF O NON-NUCLEAD RAN R METHODS

C.J. PICKFORD, J.S. HISLOP Environmental and Medical Sciences Division. UKAF.A. Atomic Energy Research Establishment. Harwell. Didcot, Oxfordshire. United Kingdom

Abstract In a laboratory analysing large numbers of dissimilar samples, of a routine ana non-routine nature, the analytical procedures usée for inor?sn:r constituents must be chosen carefully. NO single procedx;r*s is able to satisfy fully the analytical requirements for certain complex materials, and a. combination cf techniques must bt used. The quality of date retired mus weighee tb c carefully agair.s performance th t e characteristic t ir.

There is a tendency to compare analytical techniques ir s competitive fashion. Comparative list detectiof so n limits are often publishe whic) 2 , h (1 demphasis "one eth e technique against the other" approach, as well as implying that the detection limit o listecs e typical ar e achieveib n ,ca n i d any matrixconstane instrumeny ar an d r ,an fo t operator to r using that measurement principle. Performance data may also be compared in cases where separation or enrichment procedure se other havth n ei . beet caseno e n on t use ,n bu i d Manufacturers ten promoto dt e this kin competitivf do e approac e initia parth s ha f to l rus enthusiasf ho rapid man d sales growth which herald analyticaw arrivae ne sth a f o l l technique. The recent introduction cf the technique of inductively coupled plasma-mass spectrometry by Gray (3) and Houk (4) and its arrival in a commercial form illustrates this phenomenon onlwello yto . Initial sales literature implied that the technique was simple to use i.e. "black box", without interferences of a spectral or matrix nature, and could replace many existing analytical instruments directl routinn i y e operation exampler e fc ,monitorin th n i , g of water quality. Realit s show yha e techniqu nth posseso et s disadvantage wels a advantagess a l) (5 s . However? ha t i , several distinct, powerful attributes as a technique whic>. guarante analyticae placs th it en i e l armoury (6). At Harwell ,larga e numbe routinn routinf ro no d ean samples are analysed by a selection of analytical techniques, in order to establish the compositional data required for research and monitoring programmes. The principal techniques user compositionafo d l analysi thif so s kind include X-ray fluorescence (XRF), atomic absorption (AAS, flame and flameless), inductively coupled plasma-emission and mass spectrometries (ICP-ES and ICP-MS), neutron activation analysis (RNA and INAA) and ion chromatography (1C). Examination of the application of these techniques for the determinatio constituentf o n difficuln si t matrices sucs ha biologicad an h alloysas y l fl ,materials , illustratee sth complementary proceduree naturth f eo s rather thae nth comparative one. Choice of Analytical Technique When a new problem is presented to the analyst, he must mak decisioa e procedure th employede n b o n o et . Assumine gh s availablha ranga e f techniqueseo thad ,musan e th cose tb t conscious (both factors probably apply in most government and major industrial laboratories) e firs,th t questio wile nh l probably ask is: "what quality of data is required?" The chosen procedure will then probably be the cheapest one able supplo t y dat meeo at t that quality requirement. In Table 1 are listed the most frequently required levels of accuracy encountere e Chemicath n i d l Analysis Grout pa Harwell. It should be emphasised that accuracy can only be measured by the routine measurement, with samples, of SRM or M materialsCR a simila f o , r e natursamplesth o t ed an , preferably "blinde analyste absencth th o t n "I f .o e suitable reference materials, a sub-set of samples may be analyse n alternativa y db e procedure employina g fundamentally different measurement principle. Agreement shoul withie b d e limitnth s required n examplA . f thio e s approac s follows:a s i h -A stud measurin) (7 y uptake th gd an e excretio f heavno y metalrequiren ma n i se measuremen dth f to lead in blood, and arsenic in urine. The former was measured by flameles controa s a s5 lAAS 95 upo, M nSR usin S gNB accuracy. One internal reference sample was repeated every five sample precisioa s sa n estimation procedure. Arsenin ci

Tablg 1 Accuracy Requirements for Elemental Analysis

onln assessee Ca yb y analysib d f materialo s f knowso n composition . SRM eg ,(preferabl M ,CR y blind).

t majo A % 1 r relativlevels:e- accuracy required

At minor levels:- 1-5% relative accuracy required

At trace levels:- 5-25% relative accuracy required

At ultra-trace levels Any answer will do!

196 urine was measured using ICP-ES. No reference material was available, and therefore 10% of the samples were analysed by RNA. Agreemen withis t wa whic , deemes 3% ne hb wa o dt sufficiently good for the quality of analytical data required. Having decided upo e qualitnth datf yo a needen a n i d analysis, choicprocedure th f eo e used will probable yb decided by cost. It may be, of course, that only one procedure will be able to meet the requirements: this is ofte e cas ntracth t a eultra-tracd an e e levels e cosTh f .to an analysis in a commercially oriented laboratory such as our own is a composite of labour, instrumental and overhead costs: labour costs vary widely sinc e skilth e f stafo l f required for some types of analysis, for example, Group I and II element waten i s y flamrb e AAS clearls ,i y much lower than that required for accurate work near to the detection limit btechniqua y e suc ICP-Es ha r INAASo . Instrumental costs vary widely, ranging from a few pounds for some techniques, masr 1250.00o ufo t ps o spectrometerss r 0o . In general, in most laboratories in the UK today, labour mose costth te significansar t ones: thi tendes sha mako t d e rapid, instrumental techniques cheaper than labour-intensive procedures such as wet chemical methods. A qualitative indication of analytical costs in our laboratory is given in Table 2. The range of costs is approximately a factor of fifty.

Table 2 Relative Unit Cost of Analysis

Cost Technique Typ f Determinatioeo n bloon i e dS RNA ICP-MS Hf in Zr steen i o lC ICP-ES GF-AAS Pb in blood

INAA Multielement eg. Flyash

ICP-MS Multielement waters

XRF ICP-ES Multielement waters in £ l simple matrices

Flame AAS Single elements in water

Complementary Natur Analyticaf eo l Techniques In practice, if multielement analysis of a difficult matri s involvedi x e singlon , e procedur wel y t solvema no l e the problem. A good example of this is found in the analysis pulveriser o h as y ofl df fueh (pfaas l ) (8). Vast amountf so these materials are disposed of every year in the UK.

197 Because ris th f eventuao kf o e l contaminatio waterf o n e ,th CEC has set stringent limits upon the maximum levels of certain element e wastth s e leache f eb o tha y t tma ou d disposal site: accurate analysis of the matrix is thus essential. If we look at the relative sensitivities and costs o^ various methods for multielement analysis of fuel ashes for a range of toxic elements, it is apparent that ICP-ES is probably the most attractive procedure. However, some element difficule sar dissolvo t normae th y leb procedures employed otherd lose.ge b an ty .a sB durin ma g dissolution, e.g. Se, As. The sensitivity of ICP-ES is also insufficient for the ready determination of elements like U, Th, Sb, Se practicn I . andTl e therefore combinatio,a techniquef no s wil usede b l , with ICP-ES providing most elemental data, INAA being used for elements such as Sb, Se and Ba, and ICP-KS e lanthanidth d an l eT givinelements, Th , gu n datA r . fo a added bonu thas si o sete obtaine b ttw datf sy o ama r som fc de

elements, providing a valuable cross check upon acccrac.. y bctb fr C Dattechnique.d an a s obtaineA r - fo thin i dy swa 1 fcr 200 pfa samples analysed over a period of six rronthi, showec. very close agreement between the two sets of data with ? mean difference of ± 8% between data sets over the range 20 to 30c ^g/g As or Cr, The analysis of reactor grade steel sampler (9) provides another gc-cc exampl e complementarth f c e y natur* o e procedures: tho techniques used fcr multielement analyses -r these material e listear s t rr-ajoA n Tabli d . r3 e an.; irir.o. element levels. XR>" is the most cost-effective procedure tc~ the determination of most of the desired metallic components. necessary for alloy identification and «specif icat rr test:r.r XRF is capable of reaching high levels of accurar • an* precisio f suitabli n e standard e availablesar - it ; an , relatively economical s heavilsinci t i e y autor.atec.

b>j? A_NALY$IS_DF REAC

LEVEL io TECHNIQUE

Ni, Cu. Mn etc 0.01 - 100 XRF O, N, C. S 0.0005 - t-ECO Nb. Co 0.0001 - 0.01 ICP-ES Eu, Ag - 0.0005 NAA - 00005 GPAA CO ,

abovS d ^g/5 ean determinee C g ar , N Level, 0 y b df o s f Leco oe us CS14 d TC-13an 4 6 elemental analyses. Againe th , unit cost of analysis is extremely low, due to automated operation. Elementwhico C e importand ar han sb N suc ts a hfro e th m activation product point of view are present at levels below the limi f determinatioo t f XRFno : ICP-E uses i S d therefore,

198 although considerably more costly tha F whenXR n a use n i d single element, high resolution mode. Extremel w levello y s of Eu and Ag (less than 0.1 nq/q typically) which are also potential activation products measuree ,ar RNAy e higb d Th .h cos f thio t s techniqu justifies i e e th thin s i di st i cas s a e only procedure readily capable of reaching these levels of concentration. It is also, of course, blank-free which is an important pre-requisit ultra-tract a e e levels. C and 0 are also measured at very low levels by Gamma-photon Activation Analysis. Again, although costly, this techniqu abls i ereaco t e, 0 hd veran w level C lo y f o s and can discriminate against blank and surface contamination effects, which limit other, non-activation, procedures. Conclusions Analytical procedure e complementaryar s s therd a an ,s i e yet no analytical panacea. The approach adopted in a busy analytical laboratory must therefore reflect this fact, and techniques must be chosen to match samples, rather than vice versa. Rather than a competitive comparison between nuclear and non-nuclear mechods being proposed therefore, it is important to recognise that both types of procedure have their merits, and each must be applied when most appropriate.

REFERENCES ) MORRISON(1 , G.H. CriticaC ,CR l Review Analytican i s l Chemistry, (1979) 287. (2) CROSBY, N.T., Analyst 101 (1977) 225. (3) DATE, A.R. and GRAY, A.L., Analyst 108 (1983) 159. ) HOUK(4 , R.S., FASSEL, V.A. FLESCH, G.D., SVEC, K.J., GRAY, A.L., TAYLOR, C.E., Anal. Chem. 52(1980) 2283. (5) PICKFORD, C.J. and BROWN, R.M. Spectrochimica Acta, 41B (1986) 183. ) BROWN(6 , R.M., LONG S.E., PICKFORD, C.J., AERE, M3497 (1985) HMSO. ) PICKFORD(7 , C.J., LALLY publisheSHERLOCKe , b , A. o t , ,dJ. (AERE). (8) PICKFORD, C.J., MORTON, A.G. and SANDERS, T., Proc. Environmental Contamination Conference, UNEF/IRPTC (1984) 536. ) BOOTHBY(9 , R.M WILLIAMSd .an , T.M., AERE R11107 (1983) HMSO.

199 DISCUSSION

R. Schelenz. Does the number of individual determinations also include multielement methods suc s ICP-AEa h r INAAo S ?

C.J. Pickford. In my slide I included both samples and determinations for this reason e ratith abous :i o , whico betwee6 t tw he implieth n s thae th t average sampl s analyzei e r thifo ds numbe f elementso r . Mane lessar y , e.g., measuree ar c T d, Cs individuall , Th , U somn i y e leachate wated an s r samples and man y5 elements 4 sample r fo e analyze P ar s. IC y b d

200 TRACE ELEMENT ANALYSIS IN BIOLOGICAL MATERIALS COMPARISOA : NEUTROF NO N ACTIVATION ANALYSIS WITH OTHER TECHNIQUES, ESPECIALLY ATOMIC SPECTROSCOPY

P. SCHRAMEL Institu r Ökologischfü t e Chemie, Gesellschaf r Strahlenfü t Umweltforschund un - g mbH, Neuherberg, Federal Republic of Germany

Abstract

e analyticaTh l chemistry laborators concernei F GS f do y with trace element analysis mainl connection i y n with life sciences t analyseI . s annually 20.000 samples with an average of 20 elements per sample. Several different analytical techniques, including neutron activation analysis and non-nuclear method e availablear s e papeTh . r discusse e advantageth s d disadvantagean s s e differenth f o t analytical technique s appliea s o biologicat d d an l environmental samples.

Our laboratory in the GSF is concerned with trace element analytical chemistr f 1966o yd ,en sinc e especiallth e n i y the life sciences. Our main task w aresno :

1. A central analytical laboratory for trace elements in environmental samples, especiall Bavariar fo y . 2. A central analytical laboratory for the whole research center (GSF). 3. Own research programmes in the bio-medical field.

The number of samples which will be analyzed per year is about 20.000 for 20 elements on an average. samplee th e typ Th f eo s varies from different body fluids like blood serum, amneotic fluid, breast mild otheran k p su to soils, sediments and sludges. This variety describes also the variatio concentratioe th n i n n differene rangeth f so t trace elements.

201 e elemenTh t spectrum whic routinels i h y investigated covers , Na , Mn , Mg , Li , K , Hg , Fe , Cu , Cr , Co , Cd , Ca , B , Al,As . Zn d an V , Ti , Sr , Sn , Se , Sb , Pb Ni, ,P

wels Ii t l known, that there doe t exissno universa n a t l analytical technique which is able to detect this element spectru vere th y n differeni m t concentration ranges. Therefore always different analytical method necessare ar s y to solve the given problems. In general, it is impossible to discuss the capacity of on analytical method without dis- cussing the problem simultaneously. In the following I will y statementm e se s alway n relatio i sproblee th e b o o t nt m solvedr laboratoryou n .I have followine w , th e g different analytical methods available (the succession is not a jud- e chronologicath gemenvalue o t s i s t a ti , l e ordeth f o r application in our laboratory.

Neutro. 1 n activation analysis (instrumental (INAA witd an )h radiochemical separation (RNAA)) Atomi. 2 c Absorption Spectroscopy Flameles) a s AAS, graphite furnace, Zeeman-background correction b) Flameless AAS, hydride system Flameles) c s AAS, cold vapour . ICP 3 DCP-emissiod -an n spectroscopy 4. Voltammetry (DPASV)

Before I come back to the discussion of the different methods, I like to say some words to the elements in which we are interested now. e bio-medicath n I l mainle fielar e yw d interestee th n i d known essential trace elements Fe, J, Cu, Mn, Zn, Co, Mo, Se, Cr, Sn, V, F, Si and Ni in all kinds of tissues and body fluids for clearing up the role and the function of trace elements in the organism. e environmentath n I l field specific attention mus e calleb t d upon the known heavy metals like Pb, Cd, Hg, As, Tl and other more. From main interes alss interactioe i t th o f no these elements organisme witessentiae th th hn i e . on l

202 Therefore not only the pure pollution view is in the fore- ground als t biochemistre ,bu o th f theso y e elements which have a toxic action in low concentration ranges. But the mai nenvironmentae pointh n i t l controo fielt s e i d th l loading of our environment with different substances. Here, we have a lot of different matrices for analyzing purposes: soil, sediments, water, air, aquatic and terestial, plants, animals and also tissues and body fluids of man.

What is the capacity of these mentioned methods in these application fields: ad 1) In general NAA does not work as a routine method for a great numbe f sampleso r . maie th n f difficulto e On concernes i y d witcontinoue th h s decrease in number of the nuclear reactors available for this technique continuoue othen th .A s i r s increase th n i e safeguard workinr fo s g with radioactive materiae th d an l permanent denunciation of radioactivity in the public opinion, so that it is a hard task to find somebody who is willing to work in a so called hot laboratory. On the other hand the laboratory equipment becomes more and more expensive due to these safety regulations, which also influences the "price" for the analysis. But these are more political problems and do not correspond in the different countries.

From the scientific point of view NAA is a powerful tool for trace element research work, als r testinofo g other tech- certificatioe th r niquefo d san f standar no d reference materials e greates.Th t advantage completa e ar s e independence fromatrie th m x (chemical nature) e lac f th ,blano k c values, the absence of contamination problems during the analysis possibilite th , handlo t y sample th e e withouy an t pre-treatmen possibilite th d an tf o deteco t t y lo a t elements simultaneously.

But thes e theoreticaar e l advantages n practicI . e musew t problee comI th lood elementsr e man fo kth bac o t k , matrice applicatiod san n fields havI , e mentioned beforeI .

203 want to restrict me now to biological material. With INAA n solvca e onlon vera y y smal analyticae l th par f o t l problems. Due to a normally high 24 Na and 32P activity in these kind f sampleso u mussyo t work either with very short irradiation time r detectinfo s g short living isotopes hale ith nf live rang f minuteeo r witso a hlon g irradiation time including a long decay time for detecting long living isotopes firse th tn .I cas e needeon s appropriate reactor facilities like a fast rabbit system combined with a suitable detecting system very near of the reactor itself. The other method with the long living isotopes is more or less independen distance th n o t e between laboratord an y reactor. Wit INAe h th botA f techniqueo h s restrictei e son d to the elements Se, Zn, Fe, Cr and Co (sometimes Cu and Mn; it dependconcentratione th n so n biologica)i l materialr Fo . the toxic heavy metals we have normally no chance due to very low concentrations. It is therefore only a small part of the interesting element spectra. Combining NAA with radiochemical separation after irradiatio n extremlca e non y extend the number of elements depending on the chemical and time expenditure r laboratorou workinn e I .ar e w yg wite th h radiochemical separation method developed by Samsahl and we can detect up to about 40 trace elements simultaneously from vera s y i weightg m timt i e 0 samplet on 10 expensivBu . f eo e method in the chemical procedure and in Y~spectroscopy and therefor e persoon e n onlnca y sample0 1 proces o t r spe p su week d thi mucan ,s i s o les r t routinh fo s e analysise on d .An need a s"hot " laborator handlinr fo y radioactive th g e materia wele lth lwit l know al hd expensivnan e special equipments. maie th n f disadvantageo e on Thi s A i s(INA NA d an Af s o RNAA), that it is a very time and money expensive method. For diagnosis in the medical field or control in the environmental field, it is not possible to wait for the results for some weeks or months and even some days are to medicae lon th mosn i gf o tl applications (diagnosir so therapy control) interestine . th Mos f to g elemente b n sca detecte dn hig a onl hy b yexpenditur e using radiochemical

204 separation methods and therefore the costs of the analysis become very high. Adding the normally high costs for irradiation in the reactor, only some special equipped researcanalysie th y r researcpa hfo s d laboratoran ho d n ca y application on a small number of samples but never for routine analysis.

NAA was for long period the most sensitive analytical technique and most of the research work detecting the essentialit d als e toxicit an ydifferene oth th f o y t trace elements was done by using NAA-techniques in the past. From this techniqu e havew e learne importance th d f traceo e element mann i s y field d outgoinan s g from thidevelope th s - men f otheo t r analytical technique s starteswa d forcedan d .

Today for routine analysis in the wide range of applications the atomic spectroscopy method wele ar sl established. Here one must distinguish for practical work between atomic absorption (AAS) and atomic emission spectroscopy (AES). AAS was for long period after the development of the flameless techniques (graphite furnaces y thietc.b e increas th sd an ) e sensitivite th n i onle th yy competito NAA-technie th f o r - ques. Due to the development of the different plasma exci- tations, atomic emission spectroscopy has become a new peak-time since about 1978 in practice.

In general, the most considerable disadvantage of all the atomic spectroscopy methodneee f mineralizatioth o d s i s f no the samples. There exist also some newer techniques for solid sampl t plae w no n themeasurementa yno o d yo t p u t bu s important role in routine analysis due to the well known main difficultie calibrationn i s . They have their placr fo e investigating homogeneity of samples, for instance in case f standaro d reference materials norman i t bu ,l analysis there doet exissno t suc wela h l homogenized sample thaa t sample intake of 1 mg say enough about the whole sample.

205 problee Foth r mineralizatiof o m givea f no f n o samplt lo ea different ashing procedures are available for the different matrices. They all have the well known advantages and disadvantages and one must always observe the element and the matrix in choosing an adequate technique. There is not time enough to speak about those. For biological material the pressure ashing techniqu n tefloi e wit O n hHN bomb s

AAS is for many interesting elements - essential and toxic - a very sensitive analytical technique. By flame AAS one can detect especiall matrie th yd an x elementg M , Ca s, K lik, Na e P, or the essential trace elements Cu, Fe and Zn. By the flameless technique using graphite furnace and Zeeman - instead of deuterium - background correction one can detect th mose th heavt , interestinyCd d metalan b P s g essential trac f otheeo element rlo interestina td todaan e S y g elementetcl A , . Ni Finall s, Cr likr instancy, fo e Mn , Cu e one can use the hydride system or cold vapor AAS for detecting elements like As, Se, Sn, Sb and Hg. One sees there is a wide spectrum of elements detectable, but the main disadvantage is that all the AAS-techniques are single-element- methods. You have to change the light-source for each elemen u wan measuryo o tt t d thereforan e time th ee consumptio r analyzinfo n r instancfo g 0 element2 e e on n i s sampl s relativeli e y n everhighi t y bu , case lower than using NAA. An other serious disadvantage is the appearence of the different chemical matrix influences due to the relativ w temperaturlo e e availabl differene th y b e t excitation source sf systemati o whic t n causlo ca h a e c errors. The use of adequate standard reference material (matrix and element concentration range should be very near to the unknown sample s necessar)i avoio t y d these g sten bi i p.A the direction of accuracy of AAS-measurements was the introductio f Zeemao n n background correction because this

206 technique is much more specific than deuterium background correction and works sufficient over a wide range of disturbing influences.

Witnewee th h r fast data n alssystemca o e observson e each signal - the specific and the unspecific one -, therefore with some experiences one can obtain true results. In genera e instrumentth l e relativelar s y cheep,o n ther e ar e special demande laboratorth n so y equipmen d thee an tar y eas handlo t y technicae th als y ob l staff. e operatinTh g cost dependine sar e technique th n o g mose th ;t expensive one is graphite furnace AAS due to the high price for the graphite tubes and due to there low expectation of life.

r laboratorou n I y AAS-technique e usear sd especialle th r fo y determinatio f loweno r concentration elemente th f , so sPb direc, Al e determinatio, S t Mn , Ni , Cr bloon i n, Cd d serum (graphite furnace) and using the hydride system for the elements Hg, Se, As. Sn and Sb is in most of the bio-medical samples not detectable by this technique due to their low concentrations.

For our purposes we have 3 instruments combined with graphite furnace, Zeeman background correction and fast data systems and also 3 instruments combined with hydride systems. For the big amount of samples we have to analyze it is neccessar havo e t techniciay on e r eacfo nh instrument.

e otheTh r most common used atomic spectroscopy techniqun i e our laboratory is AES using plasma excitation sources. We have ICP and DCP plasma. Their is no big difference between P plasmDC y thae sa o ath t bu n n generahav i e tca edu e on l the lower temperature (about P 600 IC compare) 0K e th o t d (about 8000 K) more chemical interferences but is not possibl o intg o t edetail hige th h s o t excitatiohere e .Du n temperature in both of the plasma sources the emission line spectr e verar ay comple d thereforan x mose th et important demand is for an efficient spectrometer with a high optical

207 resolution power. Otherwise one has to fight against a lot of physical interferences, especially line coincidences f systematio t lo whic a y lea co hma t derrors n generaI . u yo l hav o tese reliabilitt e th t a result leas a f o yn a to t second emission line and/or using again adequate standard reference materials. One of the big advantages of this technique is the possibility of doing simultaneous multielement- analysis. But beginners in this technique should always start with the sequential technique using a monochromator because onl y thiyb s techniqu n collecca e e th teon experiences whic necessare har r applyinfo y simule th g - taneous method casee th mosn .I sf o tther e exist some comparable (from the sensitivity point of view) emission line f eaco s h element t comparablno t thee bu ,ar y r fo e getting true results in different matrices due to different interferences from other elements which are dependent on the different concentration ranges of the disturbing elements to the elements to be measured. By using a monochromator one e circumstanceth n tesl ca al t d choossan mos e th et sufficien ta give linr nfo e problem y usin .B simultaneoue th g s technique one must believe the computer and with no experience the results are mostly wrong. An other very importan tnecessite pointh s i t f backgrouno y d correction becaus backgroune th e e stronglb n ca d y influencey b d different salt concentrations, by different density and viscosit e samplesth f o yy differen ,b t matrices etc. Therefore an adequate background correction is necessary in both techniques - sequential and simultaneous. For sufficient short and long time stability of the instruments a good and effective gas regulating system - at best using electronic mass flow regulators at least for the aerosol necessarys i carrie - s ga r . Als peristaltioa e c th pum n i p nebulizing system shoul e installeb d a reduce r fo d d sample consumption especially for bio-medical samples in which one f informationo t smala lo f o a l wishet t samplge ou s o t se amount mostly available.

Plasma emission spectroscopy is a very sufficient tool for trace element analytical chemistry in all kinds of sample

208 material e sensitivit.Th d specifit an ymethoe th f do y allows th f einterestino detectiot lo a f gno elements alst .Bu o here one must know the limits to avoid systematic errors. Froe lis f elementth mo t s routinele whicar e hw y investigating in bio-medical and environmental samples (given at , K , Fe , Cu , Ca n detect , thca B e, P beginning:Al DC d an P )IC Li, Mg, Mn, Na, P, S, Sr, Ti, and Zn. In some cases, especially in some environmental samples like soils, sediments, sludgeV sd etcan b .P alsconcentration, e Ni th o , Cr , Cd f so are detectable. Using also a hydride system you can in these samples als , sometimeoSe detecd elementan e b th S s tA , sHg and Sn. One can not say in general which elements can be analyzed and which not, because it depends always on the concentration and on the matrix. But this is valid for all the different analytical methods.

Sufficient done plasma emission spectroscopy need sa longe r tim f experienceo e t leasscientista ,a e yea r on tfo r e .Th instrument e relativelsar y expensive, especially combination f sequentiaso d simultaneoulan s e deviceth d san operating costs are also relatively high due to a high gas (argon) consumption. Time consumption for an analysis is relatively low. We have 3 sequential and one simultaneous instrument simultaneoue th n I . s machin e havew e installe6 1 d channel differen6 1 r sfo t elements e ana on e tim-r .Th fo e lysis including three repetitions is about 4 minutes. The sample consumption is about 0.8 ml/min. Especially the new Plu8 3 sY J from Instrument Franc, vera sSA s yi e safe operating system and combined with an adequate sample changer, we can use also the night-hours for analysis withou supervisiony an t .

developmentw ne e Th P coupleIC f so d mass spectrometer promise in my opinion to be succesful in future. Today the difficultie t reallno o sd y allo applicatioe wth routinn i n e analysis.

A completely other analytical techniqu Voltammetrys i e . Ther f differeno e t exislo a tt subtechniques withine ar e .W

209 using DPASV (DiCferental Pulse Anodic Stripping Voltammetry) witmercury-droe th h p technique. This méthodmose th t s i e sensitiv thin i se comparisoeon r somfo ne elements especially Cd, Pb, Tl, Mi, Co, Cu and Zn. These are the elements whic e analyzhw e routinel y thib y s methode .Th technique expects a great deal from the quality of the ashing prodecure because small amounts of organic material disturb the analysis of all the elements due to a high backgroun peak-formd ba d a signao t .r lo

Therefore special ashing procedure r biologicafo s l materials must be applied. The normal pressure ashing technique with HNO is not sufficient enough. One must use additionally HClO r aci,o d combination havo t s e sufficient sample 4 solution r thifo s s method. Fromultia e theory th s mi -t i , e thielemenus sn advantagca t e methoon t ebu d only whee th n concentration differene th f so t element t diffeno o sd r very much. Otherwis e muson e t work sequentially e sam.s Th i e valid f.i. for the determination of Ni and Co because one must have an other electrolyte. The technique can be used for the lower ppb-range or down to the upper ppt-range. Thes w concentrationlo e s require special laboratory equipment and special chemicals for very low blanc values. The instruments are relatively cheep in comparison to the atomic spectroscopy e analysion t bu , s take relativela s y long time because all the measurements must be done by the standard additio determinatioe th n methodr fo b t P .Bu f no and Cd, sometimes also for Ni and Co in f.i. milk-powder, breast milk, blood serum, urin d othean e r body fluidss i t i , often the only method of choice.

Goin beginninge th bac o t k problee r betteo gth o o fint rt m d competitive NAA-techniquese methody th sa r o t fo s s ha e ,on that there does not exist any universal analytical method. This includes also NAA. thinI k each metho competitoa s i d r for an other method for a special given problem. The problem to be solved demands for on adequate technique. This includes all the questions about the element, the matrix, the quantit f samplo y e material availabl analysise th r e fo e th ,

210 application in practice (routine analyis or research; time consumption especially for diagnosis and therapy control) and other more. Only combination f differenso t analytical technique heln n solvinsca i p varioue th g s problem n traci s e element analytical chemistr n bio-medicai y d environan l - mental field o seriou.N s analyse t th solvo shoul t l y eal tr d problem e techniqu on s laborator y hi sb n i e s y whicb yha e h chance. This would be in most of the cases a violation of the method and should be avoided for internationally better and comparable results in trace element research and practical applications. d Plasma-AEan S AA Sn usefu a promis e b lo t combinatioe a r fo n relatively wide application field. At the end again I want applicatioe tth o neee d pointh an d o t tf wel o n l certified standard reference material l kindal f analyticao sr fo s l o reductechniquest onl e y th eywa s systematii t I . c errors n alsca oant i dhel develoo t p w methodne p n thesi s e fields.

211 THE DEVELOPMENT OF ACTIVATION ANALYSI CHINN SI A

. XINHUT A Bureau of Nuclear Fuels, Ministr f Nucleayo r Industry, Beijing, China

Abstract

Sinc e firsth e t nuclear research reacto acceleratod an r buils wa n 195ri t 8 activation analysis methods have been develope d appliean d n Chinai d . Four national activation analysis symposiums have been held. More than 200 papers on activation analysis have been published. The paper gives a review of applications. A miniature research reactor, MNSR, designed for neutron activation analysis is described.

The activation analysi unigun a s i se metho n analyticai d l chemistr f traco y e element t formi~ n importand a s an s t part of nuclear analytical chemistry. This paper will draw an outline of the development of methodology and application of activation analysis in China. Since the first experimental nuclear reacto acceleratod an r r were buil n 1958i t , Chins ha a started research work of the activation analysis. In the 60's, the activation analysis was mainly used to analyse reactor materials, nuclear fuel semiconductod an s r material n Chinai s .

ßdobd on the use of semiconductor detector and computer technique in nuclear measurement and the success of radiochemical separation procedures much progres f activatioo s n analysis ha s bee e n?0 th mad 1 n Chinai sn i ee firs Th . t symposiu n activatiomo n analysis was held in 1978. The major part of the symposium proceeding1 involves neutron activation analysis in pile. The usf neutroo e n activation analysis with preconcentratiod an n postirradiation radiochemical separatro r determinatiofo n n fourteen rare earth , elementAU d nobl , an sIr e, metalRu , Os s n Jinlii d P nd Pmeteoritan t e describear e y severab d l investigators n thiI . s symposiu ma numbe f papero r s presented the applicatio f neutroo n n activation analysi n differeni s t areas including the determination of biological, geological

213 and environmental materials. The development in methodology were also made in the fast neutron activation(FNAA) and the charged particle activation analysis(CPAA)» Determinatiof o n 0.1 pprn carbon and oxyen in silicon by using 3MeV deuteron and 29Me particl( Vo e respectively from cyclotro s reportedwa n . Such e reacheb t result y masno b d n s ca sspectrometr d vacuuan y m fusio r innero n s fusioga t n methods.

During the period f rom 1978 to the mid 1985 the activation analysis in the world was becoming a mature techique. In the same time Chin s alsha ao made great progrese th n si developmen f activatioo t n analysis technique r example.Fo , computer programs for comparative method of activation analysis have been studie d developean d n severai d l labs, e automatianth d c pneumatic transfer systems have already been set up in several research centers. Along with that the secon d thiran d d national symposiu n activatioo m n analysis were held in 1981 and 198^ respectively. In the meantime the national symposiu n applicationo m f activatioso n analysin si enviornmental science and geology was also held in 1983. The applications of neutron activation analysis in the verious fields of science and technology have expanded rapidly in recent years in China. This can be seen in the following statistical figures of the publications^~6 related to the activation analysis.

Field of Science and Technology____Number of Published works 8 5 Geolog Cosmolog/ y y Biolog Enviornmenta/ y k & l Science Methodology 41 9 1 Standar R SM / d Material Science 17 Archaeolog Forensi/ y7 c Medicine 6 20 Total

214 The publications which deal with applications of the instrumental neutron activation analysis (INAA) for multielemental determinations havs been multitudinous r exampleFo . , determination 1 microelement4 f o y INAb s bee ha A n reported-^ A .certai n e investigatornumbeth f o r s have focused their attention o n analysin rare th e geart h element n geologicai s d biologicaan l l sample s beey usinb sha nt gi show INAd n theian i nA r papers that eight to eleven rare earth elements can be analysed^»^» The methods whic e base ar hn postirradiatio o d n chemical separation or sample preconcentration neutron activation analysis(RNAA) still remain active in spite of the possibility of lossing integrit f sampleo y r instanceFo . a ,freeze-dryin g preconcentration techniqu f wateo e r sample s beeha s n use n neutroi d n activation analysi f differeno s t kind f wateso r saraples^»^ e ultratrac.Th e n geologicali s O d an u nobl,A , eAg elementf o g ~ s 10 suc s a h cosmic and biological samples are successfully analysed by RIHA**'-'. Besides determinatioe th , f traco n, Tb e , U,ThLa , ,Sm Yb, Ta and Cs in geological and environmental samples by reactor epithermal neutron activation analysis methodse ^ar described. This method reduce interferencee th s c S , Fe f so and other metal n matrixi s e mono-standarTh . d two-standaran d d neutron activation analyses3»^ are applied in determination of trace multielements in coal and fly-ash. On the basis of usin a beag f thermao m l neutron extracted fro a beam m por f teso t t reacter promp neutroy t n activation analysis technique has been studied in order to analyse composition of coal° t alsI . o reported that e proto,th n activation thin layer analysis has been adopted to study trace erosive wear of industrial machine and gun barrel^»3»

Turing this period much emphasi n improvino s beet ha spu n g the accurac f analyticao y l results e preparatioTh . w ne f o n working standard as well as the preparation and certification of the standard reference materials (SMR) have been studiedand discussed in some papers. A new mercury standard with sulfhydryl cotton matrix for "1A has been proposed^ and its advantage is to obviat lose f mercureth e coursso th f preparation o ei y n and irradiation. It is noted that a synthetic phenolic resin matrix standard which contain ? element1 s s preparewa s d an d analysed and the stability, homogeneity and the uncertainty

215 of resulted values were described in the paper1*. The national issued standard reference materials^ Include river sediment (81-01), coal fly-ashes (82 201) and peach leaves (82 301) and feochemical SRM 1SD1-8. Among them SRM GSD1-8 have been analysed by L+] labs. The analytical methods include flame and flameless AAS, spectrophotometric d ICPAESan S AE , X?F, polargraphic, NAA, gravimetric, volumetric,flame-photometri d masan c s spectrometric methods. Thre eA hav NA labe f so provide d results fo6 element3 r elemente s th whic l s 66,7i hal s f ^wito h certified values. Furthermore, the elements including As, Ce, Cs, Dy, Eu, p'd, Hf, Ho, La, Lu, Nd, Rb, Se, Sm, Ta, Tb, Tm and U ect. have been analysed by either N-1A or other methods, and it has been proved that NAA is more efficient technique than any other metho r thesfo d e elements.

n '"he -nojiiature neutron source reactor (MNSR) designed and constructe e Institutth y b d f Atomio e c Energ s commissionewa y d in the early 1984 in China and it will make the application of N4A even more popular. The neutron flux inside the irradiation tube of this MNSR is 1x101^n/cm2.s and the corresponding thermal power is 2?kw. This MNSR is equipped with two pneumatic transfer systems, one timer ranging from I1 to 999'59"» one ca^sul cutting and transfer station as well as multi-function SPAN computer e showprogramsb n n ca thaa e MNSsaft I th .s t i eR ana simple toolin nuclear analytical chemistry and it is particularly suitable for neutron activation analysis using short-lived radionuclides. During the last two years more than 10,000 samples from various fields have been analysed by using MNSR and 64 kinds of elements have been quantatively measured with sensitivitie o ppbt p .f u hundred o s m pp f o s

References

1. Selected Proceedings of the First National Symposium on Activation Analysis, Publishing Hous f Atomio e c Energy, Beijing, 1979n Chinese(i . )

2. Tian Weizhi, Atomic Energy Sei. Techn. £0 1 (1984) 113. (in Chinese)

216 e SectioTh 3.f Proceeding o n f Nationao s l Symposiu n Applicationo m f o s Activation Analysi Environmentan i s l Scienc d Geologyan e , Nucl. Tech3 . (1984) Chinesen (i . )

. 4 Selected Proceedin e Thirth df o gNationa l Symposiu n Activatioo m n Analysis, Nucl. Tech. 6 (1985). (in Chinese)

5. Chai Zhifang, W. Herr, J. Nucl. Radiochem. £ (1) (1984) 50. (in Chinese) u , InstitutJishiLi al t e , f Atomi. o e6 c Energy Annual Report, Beijing, China, 1983, 114 Chinesen .(i )

7. Li Shenzhi, "Activation Analysis in Miniature Neutron Source Reactor", Unpublished.

DISCUSSION

R. Rosenberg. e positio th WhaA compare s NA i t f o n d with other techniquen i s China?

a uniqu . Xin-huaT s i eA tracNA .d ultratrac an e e multielement analytical technique. In China more than 13 labs have worked with NAA to solve a great numbe f topico r n varioui s s field f scienco s d technologyan e , suc s geologya h , biology, environmental science R certificationSM , , superpure material, archaeology, etc. For example, 10,000 samples have been analyzed in MNSR from various e importanc e lasfieldn th o vieyearsi th e tw tf n th o wo i S s d . an e ability in resolving analytical subjects NAA stands in the same position as ICPAES. In some special cases it is even better than ICPAES, for example it can be used as a nondestructive method. But the radiation sources are limited e totae stilth ar ld lA an pricNA highe f o e r thae nonnucleath n r analytical techniques. Now NAA is not so widely and frequently used as the AAS, ICPAES, spectophotometri d electrochemicaan c l method n Chinai s .

217 NEUTRON ACTIVATION ANALYSI GEOLOGICAF SO L SAMPLE FREN SI E COMPETITIO CASA NE- HISTORY FROM FINLAND

. ROSENBERG*R . LIPPONENM , . VÄNSKL , Ä Reactor Laboratory, Technical Research Centre of Finland, Espoo, Finland

Abstract

At the Reactor Laboratory neutron activation analysis (NAA) of geological samples is performed on an analytical service basis. The expenses f thio s activit s expectei y e coveree b income th o t dy b d . Methodd an s automatic analyzers have been developed for the low cost analysis of large numbers of samples. During the period 1973-79 20,000-30,000 uranium determinations were made annually. Durin perioe th g d 1982-85 more than 10.000 samples were analyzed annually for gold and 23 other elements. The performance and cost of NAA compared with competitive methods are discussed.

1. INTRODUCTION

When analytical technique e comparear s d with each othe a number f o r performance criteria are usually applied. Those applicable to all techniques are sensitivity, accuracy, cost and capacity, the latter usually being related e costth o .t Other aspect o takt s e into accoun e speciaar t l feature f somo s e techniques. Availability, for reactor neutron activation analysis, sampling ared depthan a r surfacfo , e analysis methodo forths d n appraisaA an .s y ma l easily be subjective. There are several reasons for this. The person doing it e methon favoui on e f b o rd y becausma a specifie c instrument happene b o t s availabl o him t e e criteri e reasoth .Th y f als o mors ma n e thai e b ao eon t importan e persoth o t tn doin e comparisoth g n thae otheth n r ones. Relateo t d this is the fact that some criteria are more scientific and interesting to evaluate. Suce accuracar h d sensitivityan y . Capacit d cosan y t being less interesting to a scientist. The persons for whom those criteria are important work in laboratories doing routine work in service for geologists, steel producers and so on. They usually do not produce so many scientific papers.

Present address: International Atomic Energ yx 100 Agenc>Bo , O A-140 P , 0 Vienna. Austria

219 Another e opeapproacth e n comparison th o o t t hou o g n o woult e b d market. Sell the services, with a price covering all the costs and see how it comes out e coul .y On tha sa d t n objectiv a thi s i s e comparison e peoplTh . e who need the results can choose between different techniques. In principle e analyticath l technique which combine l criteri al mose s th tn i afavourabl e way will be the most successful one. In practice there is not one method which cover l analyticaal s l problem e besth te techniqun i waysth t .Bu e most suitable for a specific problem will be found.

At the Reactor Laboratory of the Technical Research Centre of Finland, reactor neutron activate analysis has been practised almost since the start of the reactor in 1962. Through all this time part of the work was done on analytical service basis e analysi Th .0 element4 f o smors r leswa so e s actively marketted e wor s stronglTh .wa k y sponsore e governmentth y b d n I . 1974 starte e efforth d o commercializt e neutroth e n analysis activityn I . this paper the means and outcome of this project will be described and discussed.

2. GENERAL CONSIDERATIONS

Geological researc d mineraan h l exploratio a lon s g ha ntraditio n i n Finland. Basic research is performed by the Geological Survey of Finland and a number of universities. Direct mineral exploration is performed by the Geological Survey and a few mining companies. The universities possess modest analytical laboratorie d veran sy limited fund r buyinfo s g analytical service. It is quite common for them to have joint projects with the mining companies. In this cases the samples can be analysed by the laboratories of the mining companies for no cost to the university. The Geological Survey and the mining companies have excellent analytical laboratories with modern, usually automated instrumentation.

o selT l analysi o thit s s communit t easyno s .i y

n 197I e activatio4th n analysis group comprised four professionald an s four technicians. Excellent electronica mechanicad an l l workshops were available. The irradiation were performed with the Triga MK II reactor of the institute. Samples coul e irradiateb d e rotarth n i yd specimen e rackth n i , central thimble or using two pneumatic transfer systems.

220 Until 197wore d beeth 4ha k n performe a smal n o dl scalr fo e universities and research institutes. All the work was done with care and the results were accurate sensitivitth e attained b an en ca s gooy a yb ds a d optimal timin d radiochemicaan g l separations when needed e uniTh .t coss wa t e stronglb o t e wor d th yha k o sponsoredo s higr anybodto y fo hr bu t fa o I t y. s realizewa d that accurac d sensitivitan y y coul t easilno d e improveb y d anymore. Wha s needea decreastwa s wa d n cosi e t combined wit n increasa h n i e the number of samples. The decrease in cost was attainable through rationalization of the work mainly through automation. The increase in the number of samples again required more efficient marketing.

n 197o differeni tw 4o S t projects were startedo built n e a d On . automatic uranium analyze d anotheo builan t r n automatie a don r c gamma spectrometer. In the same time a realistic price list was worked out and the marketing was increased. The folowing face started in 1981 and still continues. The NAA was completely commercialized and the automation of the laboratory strongly increased.

3. ANALYSI URANIUF SO NEUTROY B M N ACTIVATIO D DELAYENAN D NEUTRON COUNTING

e uraniuTh m analysis projec s startee timtwa th en i whede th n exploration for uranium was intensified in Finland. In the end of 1973 a manual uranium analyzer was taken into use. Because of the low cost of the analysis it was not difficult to get customers. The mining companies sent control sample r analysie resultfo sth d an ss proved thawore th tk coule b d e Researcth don t a e h Laboratory t worthwhilno w s price. lo wa Witt e i esth h minine th r gfo companie o buil t sn analysiow d s capacity e onlTh .y probles wa m the capacity. With the manual analyzer the annual number of samples which coul e analyzee beginninb d th n s s decidei 8000 wa wa df o 197 o S gt . i o 4 t d buil n automatia d c analyser.

The project started during 1974. The analyzer was taken into use in August 1975.

The operating principles of the analyzer can be seen in FIG. 1. The device comprises a sample changer, a pneumatic transfer system, a measurement station with detector d auxiliaran s y electronics microcomputera , weighina , g systea containe e storagd e analyzeth an mth r f fo o er d samples. [1].

221 __^, , IRPAÜ ÄTiüN

SAMPLE OUT

FIG. 1. Principles of the automatic uranium analyzer.

The measurement station comprises a moderator of polyethylene in which 3 12 He detectors are imbedded in a circular configuration around the sample position. The moderator is surrounded by a sheet of cadmium and borated paraffin.

e systeTh me followinworkth n i s g way e sample.Th e loadear s d inte th o sample changer. The first sample is a blank, the next a standard and the following samples with blank d controlan s s accordin o neede t analyzeg Th . r irradiates and measures the samples automatically and shoots them into the waste container e sampleTh . e alsar so weighed automatically e computeTh . r calculates the elemental concentrations and prints the results on paper and paper tape.

n efficienA t capacit 0 samplee reached40 b n f r workino yca pe s y . da g The detection limit is 0.1 ppm. The system uses rabbits of polyethylene. Thes e transportear e boxen i d f styroo s x holdin 0 rabbit9 g s each. Thee ar y sent to the customers who insert the samples into the rabbits, mark them with the appropriate code numbers and send them back. The results are sent as paper listing r papeo s r tape.

During 197 6o 197t e annua9th l numbe f sampleo r s analyzed varied between 20,000 and 30,000. Then geocheraical exploration for uranium decreased strongly, the annual number of samples being «,000 in 1985. During the best year the income from uranium analysis was about 75% of the total income from analytical service by activation analysis, but it still covered only about

222 e e budgeactivatioth th hal f f o fo t n analysis group e uraniuTh . m analysis itself mad considerabla e e profit becaus n yearema onls2 1. wery e needer fo d the work. This time includes service of the analyzers.

4. INSTRUMENTAL NEUTRON ACTIVATION ANALYSIS

1 Analysi4. usiny b s g thermal neutrons

INAA of 31 elements [2] was performed on a service basis. Table I shows the elements and the obtained detection limits. All the work was made manually, the spectra were recorded on paper tape and the results calculated ba separaty e computer e numbeTh . f sampleo r s analyzed annually s abouwa t 3000 in small series. As a consequence the unit price was high. In order to improve the situation it was decided to apply automation. In 1975 a sample change s builwa r t [3]. Thumeasuremente th s s coul performee b d d automatically. Manual work was saved and the capacity of the analyzed increased because also the nights and week-ends could be efficiently utilized for measurements. The spectra were recorded on magnetic tape and a separate computer was used for calculations. The price of the service could be decreased somewha t competitivet stilno tbu e cos s th lwa t r institutionFo . s with their own laboratories there is always a high threshold for sending samples outside for analysis. Mainly lanthanoids and some other elements, not easily analyzed by other techniques were determined in small quantity.

Tabl . I e Detection limit1 element3 f o s rockn i s s determiney b d instrumental neutron activation analysis.

Element Detection limit Element Detection limit (ppro) (ppm)

Na 90 Ba 160 Al 300 La 0.5 K 3000 Ce 2 Ca 105 Kd 5 Sc 0.06 Sm 0.06 Cr 5 Eu 0.05 Ti 6000 Gd 4 V 30 Tb 0.05 Mn 10 Dy 0.3 Fe 350 Yb 0.07 Ni 15 Lu 0.2 Co 0.3 Hf 0.6 Rb 20 Ta 0.02 Zr 200 Th 0.4 Sb 0.2 U 0.5 Cs 0.4

223 2 Analysi4. y usinb s g epithermal neutrons

In the beginning of the 1981 there was an increasing demand for the analysis of gold in geochemical samples. Geochemical exploration of gold requires a low limit of detection so that all anomalies can be found. It also requires a low unit cost of the analysis because large numbers of samples have to be analyzed. The laboratories of the mining companies could not satisfy e geologiste needth th f o s markea o thers s s r outsidewa fo t e services wa t I . decided that the Reactor Laboratory should try to utilize the situation.

Two problems had to be solved. How to do the analysis in order to w detectiolo reac e th h n limit o organizst requirew wore ho orden th ed i k an d r o react a high h enough capacitw enouglo d a costh an y .

e detectioTh n limi 0 ppb1 t , obtainablw e lo wit t hno thermas i A NA l enough to enable the detection of all anomalies. In order to increase the attraction of the service a number of other elements should be analyzed simultaneously with gold. In Finland a very important one is molybdenum which cannot be analyzed with thermal NAA. These two aspects compelled the use of epithermal NAA. Nobod d previouslha y f ENAyo n suc reportei e A a hus e th d large scal d therefora challengean e s wa t i e .

First the method for the irradiation of the samples in cadmium had to be developed. After testing some different approaches the following was arrived at. Twenty containers of aluminium, $ 30mm x 200tnm lined with 1mm of cadmiu d 0.2man m f aluminiumo m e rotatin th e kep ar n ,i t g specimee n th rac f o k reactor. The containers are kept in the reactor all the time. They are only lifted up when samples are inserted and removed. Because of the special shielding syste e dosth m e obtaine e handleth y s negligibleb di r 0 0.564 g. samples can be irradiated simultaneously.

The samples are weighed into capsules of polyethylene which are wrapped int a thio n foi f aluminiumo l e bundleTh . e insertear s d inte irradiatioth o n containers together with standards. Thee irradiate ar ye wee on k r (35h)fo d , let decay for four days and measured for 20 min per sample. Thus the measuremen 0 e seriesample16 on f f o o ts s take 5 day2. s s usin a samplg e changer. Tabl I showI e e elementth s s detecte d theian d r detection limitn i s rocks [4J.

224 Table II. Detection limits for 25 elements in rocks determined by instrumental epithermal neutron activation analysis of till. Sample e measurear s d four days after irradiation.

Element Detection limit Element Detection limit (ppm) (ppm)

Na 250 Sb 0.1 Sc 0.5 Cs 0.6 Cr 40 Ba 80 Fe 2500 La 1.5 Co 2.5 Sm 0.05 Ni 40 Eu 2 Zn 100 Lu 0.05 tit 1 Ta 0.5 Br 0.6 W 2 Rb 15 Au 0.003 No 1.5 Tb 0.4 AS 3 U 0.3 Sn 100

Usually the samples are weighed by the customers into capsules provided e Reactobth y r Laboratory e capsuleTh . e transportear s d storean d boxen i d s of styrox. The boxes hold 100 capsules. On arrival to the Laboratory the samples are subject to the Laboratory's book-keeping system and included in the routine flow of samples. Then the samples are irradiated, let decay and measured with one of the five automatic gamma spectrometers of the NAA group(s).

AUDIO C-CASETTE UNIT (PROGRAMS]

DIGITAL C-CASETTE UNIT (DATA)

SAMPLE CHANGER Pb-SHIELD

PRINTER

DETECTOR

FIG. 2. Principles of the automatic -^-spectrometer.

FIG. 2 shows the structure of the spectrometer. It comprises

sample changer detector with auxiliary electronics multichannel analyzer microcomputer input/output devices

225 Sample changers for 66, 80 and 120 samples have been built. 0.5 - 5 ml capsule e usedb n .ca s

Four Ge-detectors with 10-25% relative efficiency and one low-energy photon detecto energn a d ran y witresolutiom m h0 0 1 dimension i 60 f f x o n 5 f o s e usedar .V ke 2 12 t ea V

The central board of AIM 65 by Rockwell is used as basis for the computer. Memory extensions, the power unit, the case, interfaces and the programs are made at the Reactor Laboratory. The I/O devices are an Epson MX-80 printer, an MFE 2500 digital cassette terminal and a philips NE2235 audio recorder.

The system functions as follows. The samples with standards first are loaded in the sample changer. The data containing measurement parameters and sample code d weightan s e loadear s d fro a cassettm e inte computeth o r memory. Thee systeth n m automatically measure e sampleth s d calculatean s e resultth d s giving as output elemental concentrations.

The elemental concentrations are printed on paper and also on a digital cassette. This can be used for the transfer of data to another computer for data processing e useth rf I doet .hav no sa cassett e e terminal e dats th ,i a transferred from the cassette to a floppy disc or computer compatible magnetic tape using other computer e instituteth n i s .

The number of samples analyzed annually by ENAA has been more than 10,000 for several years.

5. DISCUSSION

The basic starting point is that a geologist has samples he want to be analyzed and then some money. Now the first point is to find a laboratory wit e worktechniqua h th o analyz T .o d e specifio t th e c elements i e h s interested in. The requirements are the following:

e e techniquaccuratb Th o t s ee ha result e enougth e usefub r o t fo sh l and sensitive enough for the elements to be detected.

There are some differences in the accuracy of anaytical techniques but when up to date instrumentation is available, these differences are not significant. Much more importan e persone skilth th f s o li s t doin e workth g .

226 In Finland the laboratories arc of high quality and there are not significant difference n accuraci s y betwee e laboratoriesth n .

The sensitivity is important. NAA in general is a very sensitive technique when chemical separation e appliedb o n t to thaca ss Bu i .t expensive. INAA on the other hand is not especially sensitive for most elements. Some techniques are sensitive for other elements and some others r othefo r a consequencones s A . e techniquth e choses i e n accordine th o t g proble t hand a mr instanc Fo . w etechnique fe ther e ar e s capabl f detectino e g lanthanoid n theii s r background concentrations. Therefore INA mainls i A y used. The same stands for some of the platinum group metals. For most other elements at least a few different techniques are possible. Another thing is that some techniques are sensitive to concentration but they require a large sample n INAI e sampl.th A e varieb sizn ca de considerably without losinn i g sensitivity e ablb o analyz t eo T . 0 elements4 e g samplm 0 s 1 i e fro e on m important sometimes. Likewise it is sometimes important that the method is n destructiveno n geologI . y these feature e needear s o seldos d ms i tha t i t not a basis for an economical success of the technique.

Sometime e multielementh s t capabilite th f mentiones o i y e on s a d advantage f INAAo sw ther No . e exis a numbet f otheo r r techniques possessing the same feature. And then this is really a feature which only affects the cost of the analysis. A single element method which is capable of ten single element determinations in a certain time is as good as a method which is capable of one ten element determination in the same time.

One of the real drawbacks of NAA is that it can be used only in a few places becaus researca e h reacto neededs i r t wheBu .n analytical servics i e discussed this doet applno s y becaus e custometh e r always hav o sent e d their samples somewhere when thee outsidus y e laboratories.

The capacity is usually not a problem in itself. It is basically alway mattea s f costo r . Accordin r calculationou o t g e irradiatioth s n capacity in the Reactor Laboratory using 35h irradiations, would be 90.000 samples per annum if the reactor were run in three shifts. Gamma spectrometer e purchaseb n ca s d accordin o needt e gsam Th .e effect coule b d arrive t wite a currend th h t running schedul y decreasinb e e irradiatioth g n time and increase, the measurement time corresponding.

227 Table III gives examples of the cost of analytical service in Finland. It can be seen that single element (U) or difficult multielement (REE) NAA is expensive but so is single element analysis using other techniques (Au). Routine multielement analysi ss competitiv i usin A NA g pricn i e e wity othean h r technique.

Table III. Cos f analyticato l sevic f geologicao e l sample Finlandn i s . Data from price list f differeno s t companies. DNAA: neutron activation analysis and delayed neutron counting INAA: instrumental reactor neutron activation analysis. ENAA: instrumental epithermal neutron activation analysis XRF: wave dispersive X-ray fluorescence analysis. : atomiAA c absorption spectrophotometry P -inductivelIC . y coupled plasma emission spectrography. FAA - furnace atomic absorption spectrophotometry.

cost of cosf o t batch of Elements cost per Method first additiona 0 sample10 l s , determined element in sample sample cost per batcf o h sample 100 samples S * * *

DNAA 163 3.2 4.8 U 4.8 ENAA 124 16 17.1 4 7 2 4. 0 elements INAA 194 49 50.5 10 REE-elements 5.0 XRF 34 34 34 2 21. 8 elements AA/ICP 10.6 10.6 10.6 11 elements 0.96 FAA 10 10 10 Au 10

The NAA services most sold are DNAA of uranium, ENAA of 24 elements and the analysi 0 lanthanoids1 f o s . These three cases wil e discusseb l e th n i d following.

When the geochemical exploration of uranium started in Finland it immediately got considerable funding for field work. A number of samples were collecte e mininth y b dg companie t hav no t a techniqu theed bu sdi y e readr fo y low-cost sensitive uranium analysis. For XRF the detection limit is 10 ppm while a detection limit of 1 ppm or better was needed. The Reactor Laboratory was immediately ready to receive samples. Therefore the samples were sent to the Reactor Laboratory, instead of building up own capacity, which would have taken time. Probably the work could have been done in the laboratories of the mining companie y fluorescenceb s a lowert leas a r r fo ,to , comparable cost. This was never done, but the number of analyzed samples decreased, because of a decrease in exploration for uranium in the country.

228 The situation was more or less the same when the geochemical exploratio r golfo nd starte largea n i d r scale e ReactoTh . r Laboratory could analyze gold with a detection limit of 3 ppb which was barely enough for geochemical exploration. Several geologists complained that a detection limit woulb pp e needed ob d1 f e capacitTh . y coul e rapidlb d y increaseo t d thousand f sampleo s s annually. Therefor minine th mos f go t companiee th d an s Geological Survey sent these samplee Researcth o t s h Laboratory e cosr Th pe t. sample was quite high but a number of additional elements were obtained. Many of these elements were not really needed and the Geological Survey and some of the mining companies developed FAA techniques for the analysis of gold in their own laboratories. A detection limit below 1 ppb could be obtained. The Reactor Laboratory stil e samplelth getl al s from some companiese th d an , samples which exceed the capacity of the laboratories, from the Geological Survey and some other laboratories.

In Finland there still does not exist a competitive method for the analysis of lanthanoids. Therefore all the samples are sent to the Reactor Laboratory. Because this wor mainls i k y relate o basit d c researc e numbeth h r of sample s lowi s , below 1000 annually. Therefor economicas it e l significance e Reactoth r fo r Laborator onls i y y marginal.

During 1982-8 e annuath 5 l numbe f analyzeo r d sample s abouwa s t 20.000 corresponding to 300.000 element determinations. The average annual income from analytical servic s US$220.000wa e . This covere e activit th e cos f th do t y including development work. Now the trend is decreasing.

e presenTh t situatio e summariseb n ca n s followsa d : Research groups not possessing their own laboratories still send their samples for NAA because the technique is better than other ones for uranium and lanthanoids. The ENAA analysis give a bettes r picture concentratioth f o e f economicallo n y significant trace elements thay othean n r single technique t theBu y. also sen e samd TCP-analysisth dan e F sampleXR r y fo combininsB . g these three techniques a ,geochemica l e samplalmosb n ca et completely characteriseo t s a d its chemical composition.

Well equipped laboratories e otheth n ro , hand, hav a numbee f o r different techniques availabl o themt r eeac Fo .h element, there exist a t least one non activation analysis technique which is at least as good as INAA. Therefor e Geologicath e mining bi le gSurve th companie d o an yt y tr s analyze thei n samplesow r , when they have capacity enough a tim n eI .whe e th n

229 laboratories were fully occupied with work, the introduction of new capacity and methods could be difficult and outside contractors were used. But now the mineral exploration is decreasing and the instruments and personnel has to be utilized even for work which could be more economically performed outside.

n ordeI o survivt r e activatioth e n e furtheb analysi o t s r ha s developed. The selection of elements has to be increased to include new ones which cannot be easily analysed by other techniques and the cost has to be decreased. Currentl w completelne a y y automatic activation analyze r shorfo r t lived nuclide s beini s g constructed A centra. l compute alss i r o being connected with the automatic gamma spectrometers in order to obtain a centralized data handling, storage and reporting system. All this is done to further decrease the cost of the analysis but also to increase the number of different element e analyzedb o t s .

References

[1] Rosenberg, R.J., Pitkänen, V., Sorsa, A., J. Radioanal. Chem. 37(1977) 169.

] [2 Rosenberg, R.J., Suotnen Kemistilehti, 45JH1972) 399.

] [3 Rosenberg, R.J. d Zilliacus,an Radioanal. J , ,R. , Chem. 39(1977) 189.

[4] Rosenberg, R.J., Kaistila, M., and Zilliacus, R., J. Radioanal. Chem. 21(1982) 419.

15] Vänskä, L., Rosenberg, R.J., and Pitkänen, V., Nuclear Instruments and Methods 213(1983) 343.

DISCUSSION

F.F. Dyer. What is the counting efficiency of your He-3 neutron counting?

R. Rosenberg. It has not been measured but we get 100 counts per microgram U using our standard procedure.

230 SUMMARY AND CONCLUSIONS

Formal presentation M consisteAG e th f fouo t da sr main sessions:

. Researc1 h Trends, Presen d Futuran t e . Measuremen2 t systems 3. Tracers, and . Activatio4 n techniques.

This summary will address first the principal ideas developed at each of the above sessions. Information is drawn from comments at the end of each presentation, from the discussion that was held between individual session participants only d fro an ,n M opedelegates a mAG n l forual s mighf A o m. e b t expected, several important observation t acroscu se thematith s c boundarief o s a single session. When this occurred e ite th ,s discusse i m d wher t seemi e s most appropriate.

The development of analytical measurements science, in a country involves education and coordination. Such activities run the gamut from basic research on one extreme through applied research (task solving), development, production, and training. As nuclear and non-nuclear analytical technology reaches new levels of sophistication, it becomes increasingly more difficult to match critical measurement problems wite mosth ht appropriate analytical methodology. This is especially true in the developing nations where facilities, equipment d trainean , d manpowe e limitedb y rma . Thus, complex issues mus e resolveb t d before decision madee b n e .country' ca sth Wha e ar t s special needs? Are these needs located in a university, industrial, or government laboratory environment w sophisticateHo ? e locath s li d scientific infra-structure wits i ht I thes? e kind f questiono s s tha a countrt y must plan and operate its programs.

While nuclear techniques are clearly being challenged by other analytical methods, virtuall l participantal y s agree that they still plan a y importan e tarsena th rol n i f eanalyticao l l chemical science n manI . y instances this role is seen as complimentary rather than competitive. For example, it was noted that nuclear methods, when used, were primarily associated with research studies rather than extensive, routine applications. Nuclear technique n ofteca s n provide advantages that simply canno e realizeb t d

231 with competing procedures. Superior accuracy for trace element analysis or freedom from certain matrix effect e notablar s e examples a fundamenta n O . l level, however e choic a th ,specifi f o e c measurement method still dependn o s the special characteristics of the sample combined with the analytical criteria of the required results. Unfortunately, many of these basic criteria (e.g. sensitivity, limit of detection, elements of cost) are often confusing, especially when applie o different d t analytical methods.

A relatively new discipline, scientometrics, was discussed. The full potentia f scientometrico l s onli w sbein no y g recognized. Scientometricn ca s be used to indicate significant trends in research and may be able to differentiate between quality and quantity in publications and information y alsma flow ot I y suggestha. wa te analyticath t l sample e effectivelar s y linked with appropriate analytical methods t scientometriBu . c methods must stil e combineb l d with other rating techniques suc s peea h r revie d experan w t adjudicatio o achievt n a valide , broadly-based evaluation. Regardlese th f o s tactics used l sucal , h appraisal e besar s t carrie t usinou d g rigorous definition e figureth e economicr th f meri o fo e stechniqued th an tf o s s being considered. Onl f thi i y dons n i inter-methos ca e d comparison e madb s e successfully. As noted above, however, existing definitions are often poorly developed and frequently cannot serve as an effective basis of method selection. Clearly e creatioth , f preciso n e criteri r analyticafo a l methods appraisal is needed. Chemometrics, another advanced concept, may offer new and different insights for the study of analytical methods research.

A significant trend in the evolution of analytical measurement systems is the use of micro computers (PC's) as the major component of instrument control. Previously, small computers were used primaril r datfo ya collection and reduction and were often dedicated to a specific measurement application. Now, however, many special plug-in "boards" are being manufactured for personal computers that allow the unit to be configured into many different instrument a softwarvi s e control n effectI . e computeth , r become e instrumentth s . Some commercially available boards already provide functions for a multi-channel analyzer, an optical spectrum analyzer, digital and analog oscilloscopes, and multi-purpose input/output control. Thus, a personal computer, using selected software e quicklb n ca y, re-configured into virtuall y typ f an analyticayo e l instrumentation that migh e neededb t e On . decisiv f suco ehe advantagus board e greatls th i s o t e y reduced cost, especially in multipurpose applications. In the very near future, it is

232 probable that PC instruments will form the backbone of many measurement devices that are now supplied commercially as hard-wired units.

Tracer techniques represen a vert y useful are f nucleao a r analytical chemistry. Isotope dilution analysis, with the aid of both active and stable isotopes d immunoassaan , y technique e perhapar s mose th st important. Isotope dilution analysi e useb dn ca wits h biologica d geologicaan l l materials, water, soils, semiconductors, etc. Ofte e detectioth n n limits equa r exceeo l d that of other techniques while selectivity is fair to excellent. The principal and decisive advantag f isotopo e e dilutio e possibilitth s i n f usino y g non-quantitative isolating procedure o tha s sn som i t e instancese th s i t i , only analytical method that can be used to solve the problem.

Owin o simplicityt g , sensitivity d selectivityan , , immunoassay methods are becoming a major approach to the solution of many analytical problems. These techniques, though usually associated with clinical applications, are finding new uses outside this area. In the near future, the extensio f immunoassayo n s expectei s e n nephrologfieli dth f o n di d an y prophylaxis and control of cancer treatment.

Until now e reaction,th f isotopo s e exchange have been used primarily for research studies of chemical system dynamics. Reaction mechanisms, kinetics, complex formation d ioni d moleculaan , an c r mobility havl beeal en addressed. Exchange method e alsar so user radiochemicafo d l isolatiof o n e determinatioth material r fo d an sf micr o n o quantitie f elementso s . Mass spectrometry is another versatile and powerful technique capable of being employe r botfo d h organi d inorganian c c problem varieta n i s f contextso y . Althoug e formath h l presentation emphasized safeguards applicationss i t i , equally useful in the environmental, nutritional, and materials science areas. It is, however, equipment intensive and requires very capable technicians.

Future developmen f traceo t r techniques d includan w e designe th e f o n more effective reagents for isotope dilution and immunoassay, development of novel separation procedures, fashioning of techniques for the simultaneous determination of two or more analytes, and automation of the processes for routine analyses. Radiotracer techniques are extremely useful in both the develope d developinan d g countries becaus f theio e r simplicityw lo , instrumentation costs d relativelan , y short analysis time. Alsoe th ,

233 precision and accuracy of these methods compares favorably with other analytical techniques.

Activation analysis is a major nuclear technique that has significant advantages and recognized limitations. Accuracy, matrix independence, little o sampln r o e preparation, multielement capability, sensitivityd an , reliabilit y benefitske e ar y A .critica l constrain o broadet r application, however e necessitth s i , r acces a reactorfo y o t s ; these facilitie e oftear s n sparsely distribute r non-existeno d mann i t y countries. Another disadvantage of the technique is that a variety of radioactive materials are generated; this leads to the associated problems of waste handling and disposal as well as other concerns of public health.

The cost of activation methods is widely perceived to be higher than competing methods. For large-scale analytical projects this need not be true, especiall e measurementh f i y t procedur s properli e y e precisioscaleth o t d n e datath f .o e Commerciaus require e th n i ld activation analysis softwars ha e been availabl r lesfo e s tha decadea n d evean , n now n integratea , d packagr fo e irradiation, counting, and data reduction is not readily available. Costs for equipment are dropping, however. For the price of a simple multichannel pulse-height w yearanalyzew obtaife no e wit sa n on nf agohca o r bettee w , r performanc d interfacean e a compute o t d r datfo r a handling (see above)e Th . price/performance ratio of germanium detectors is also going down.

Modern method f instrumentao s l analysis frequently require major support functions suc s liquia h d nitrogen, specialty gases r conditioningai , , d technicaan d maintenancan l e staff. Activation analysi o exception s i s n and, r thifo se bes b o reasonconcentrat t ty ma t i , e new, sophisticated analytical installation t nationaa s l nuclear center r otheo s r locations where this support structure is already well developed. Modern activation analysis equipmen n existine addea b o n t d ca t g reacto comparativela r fo r y modest cost .e justifieb Thi e demany th ma s r analysif i fo d s highi s . Other approaches, such as the use of neutron generators, particle accelerators, and photon sources can be employed when special analytical requirements exist.

As noted above, the selection of an analytical method is strongly influence e typ th f samplo ey b d e being e characteristicstudieth d an d e th f o s required results. Powerful optical techniques like ICP-E n ofteca Se b n employed where activation methods were once used. Such method e clearlar s y

234 indicated whe o reacton r neutron sourc s availablei e . However, activation analysi s stili s l perceive s holdina d a commanding g position when sensitive, reliable multi-elemend an , t respons s neededi e . Thus r somfo , e applications, travel to a reactor activation facility may be both appropriate and cost effective when one is not locally accessible.

235 LIS PARTICIPANTF TO S

. C.RDr . Beverly Martin Marietta 141x P.OBo .0 Paducah 4200Y K , 1 U.S.A.

Dr. Richard W. Bild Analytical Chemistry Div. 1821 Sandia National Laboratories Albuquerque, New Mexico 87185 U.S.A.

Professo . BraurT n Institute of Inorganic and Analytical Chemistry . Eb'tvoL s University P.O.Bo3 12 x H-1443 Budapest Hungary

Dr. C. Burtis Oak Ridge National Laboratory Post Office Box X k RidgeOa , Tennessee 37831 U.S.A.

. F.FDr . Dyer Oak Ridge National Laboratory Post Office Box X k RidgeOa , Tennessee 37831 U.S.A.

. J.SDr . Eldridge Oak Ridge National Laboratory Post OfficX x Bo e k RidgeOa , Tennessee 37831 U.S.A.

. MariDr a Guecheva Swiss Federal Institut f Forestro e y Research CH-8903 Birmensdorf/Ziirich Switzerland

Dr. E.L. Huffman McMaster University Hamilton, 1 Ontari4K S L8 o Canada

Professor Nagao Ikeda Department of Chemistry The Universit f Tsukubo y a Sakura-Mura, Ibaraki, 300-31 Japan

237 . J.RDr . Jewett Rockwell Hanford Co. 0 80 x P.OBo . Richland, WA 99352 USA

Professor E.H. Klehr The University of Oklahoma 202 West Boyd Street Norman, Oklahoma 73069 U.S.A.

Professor V. Krivan Universität Ulm Sektion Analytik und Höchstreinigung Postfach 4066 D-790m 0Ul Federal Republi f Germano c y

W.R. Dr . Laing k RidgOa e National Laboratory P.O. Box X, k RidgeOa N 3783T , 1 USA

. RicharDr . LindstroM d m Centre for Analytical Chemistry United States Department of Commerce National Bureau of Standards Gaithersburg, Maryland 20899 U.S.A.

W.S. Dr . Lyon Oak Ridge National Laboratory Post OfficX x Bo e Oak Ridge, Tennessee 37831 U.S.A.

Dr. R.L. Mayer Safeguards R+D Monsanto Research Corporation P.O.Box 32 Miamisbourg, Ohio 45342 U.S.A.

Dr. John McCown Westinghouse Hanford 197x P.OBo .0 Richland, WA 99352 USA

Dr. F.R. Mynatt k RidgOa e National Laboratory P.O. Box X k RidgeOa N 3783T , 1 USA

238 Dr. C.J. Pickford AERE Harwell Oxfordshire, 0X1A 1OR United Kingdom

Pruety Ro . tDr k RidgOa e National Laboratory X x P.OBo . Oak Ridge, TN 37831 USA

. DaviDr . RakeA d l Safeguards R4D Monsanto Research Corporation P.O.Bo2 3 x Miamisbourg, Ohio 45342 U.S.A.

. ReveG . lDr Laboratoire d'Analyse par Activation Pierre Sue Centre d'Etudes Nucléaires de Saclay B.P. no. 2 F-91190 Gif-sur-Yvette France

. RosenberR . Dr g (Scientific Secretary) IAEA P.O.Bo0 10 x A-1400 Vienna

Dr. H.H. Ross k RidgOa e National Laboratory Post OfficX x Bo e k RidgeOa , Tennessee 37831 U.S.A.

Dr. R. Schelenz IAEA P.O.Bo0 10 x A-1400 Vienna

Dr. P. Schramel Gesellschaft für Strahlen- und Umweltforschung mbH (GSF München) Institut für Ökologische Chemie Ingolstädter Landstrassl e D-8042 Neuherberg Federal Republic of Germany

. ToshiDr o Shigematsu Ibaraki Electrical Communication Laboratories NTT Tokai-Mura, Naka-Gun, Ibaraki-ken 319-11 Japan

239 Dr. W.D. Shults k RidgOa e National Laboratory Post OfficX x Bo e Oak Ridge, Tennessee 37831 U.S.A.

Dr. D. Smith k RidgOa e National Laboratory Post OfficX x Bo e Oak Ridge, Tennessee 37831 U.S.A.

Professor J. Tölgyessy Department of Environmental Chemistry and Technology Slovak Technical University Jânsk1 . aUl 81237 Bratislava CSSR

Ing. Tian Xin-hua Burea f Nucleao u r Fuels P.O.Box 2102(10) Beijing People's Republi f Chino c a

240