Journal of Radioanaly tical Chemistry, Voi. 37 [1977) 801-811

DETERMINATION OF TRACE ELEMENTS BY NEUTRON ACTIVATION ANALYSIS USING DINONYLNAPHTHALENE SULFONIC ACID AS A PRECONCENTRATING AGENT M. H. YANG,* P. Y. CHEN, C. L. TSENG, S. J. YEH, P. S. WENG Institute o f Nuclear Science, National Tsing Hwa University, Hsinchu, Taiwan 300 (Republic o f ChinaJ (Received November 15, 1976) Dinonylnaphthalene sulfonic acid (HD) has been used as a preconcentrating agent to enrich trace metal ions and to separate the interfering elements such as Na, K, CI and Br which normally exist in the natural aqueous systems. Experiments were performed by extracting the ions in the aqueous medium with HD in n-hexane and subsequently backextracted into a minimal volume of acid solution. Factors influencing the extraction efficiency of the ions of interest were investigated. The preconcentration technique developed in this study was applied to the determination of trace elements in biological and natural water samples by neutron activation analysis.

Introduction The preconcentration o f trace elements has assumed great importance not only in the field o f high-purity materials but also in connection with environment. Though radioactivation analysis is well known for its high sensitivity, the fact that the levels o f certain heavy metals in biological or other natural systems are found to be t o o low to be determined directly, implies the necessity for the concentration prior to the measurement. Moreover, the major components in the matrix substances, e.g. Na, K, C1 and Br which have relatively high thermal neutron cross-sections, may become serious interferences in the gamma measurement o f trace components. Thus, the enrichment o f trace elements o f interest and the separation o f the major interfering components constitute the main interest in the preconcentration in neutron activation analysis. Great care is, however, necessary to avoid contamination as well as loss o f determinant at all stages o f the preconcentration procedures. *Present address: Max-Planck-lnstitut for Metallforschung Stuttgart, lnstitut f'or Werkstoffwissenschaften, Laboratorium fiir Reinstoffe, Schw~ibiseh Gmiid, FRG. J. Radioanal. Chem. 37(1977) 20*

801

M. H. YANG et al.: DETERMINATION OF TRACE ELEMENTS

Solvent extraction has been widely used for the preconcentration of trace elements in neutron activation analysis. ~-2 A number of organic extractants, such as dithizone and oxine, 3-4 have been used to extract a group of metals which were then differentiated and determined by activation analysis. Dinonylnaphthalene sulfonic acid (hereafter denoted as HD) is known as a liquid cation exchanger which forms micelles in nonpolar organic solvents s-7 and exchanges with a metal ion M b§ by the following reaction MbA~ + (HD)x, o = (MHx_ b Dx)o + bH+A

(1)

where X is the number of aggregation of a HD micell and the subscripts A and 0 denote the aqueous and organic phases respectively. Since HD gives stronger extraction of metal ions from the aqueous phase to the organic phase by a factor of more than 103 , in comparison with the same formal concentration of the chelating extractant, thenoyitrifluoroacetylacetone,8 we have attempted to investigate the feasibility by using HD as a preconcentrating agent in neutron activation analysis. The preconcentration of trace elements was accomplished by extracting the ions in aqueous medium with HD in n-hexane and subsequently back-extracted into a minimal volume of acid solution. Attempts were made to find an optimum condition to eliminate the interfering elements (Na, K, Br, CI) which normally exist in the matrices without sacrificing the extraction efficiency for most of the metal ions.

Experimental

Materials and equipments HD was obtained as a 30% solution in n-heptane from R. T. Vanderbilt Co., Inc., U. S. A. The materials were further purified by repeated extraction with 3N HNO3. All reagents used (HNO3, HC104 and n-hexane) were of guaranted high-purity grade (Suprapur grade from E. Merck Co.). Redistilled water prepared with a quartz distilled apparatus was used throughout. Radioactivity measurements in the activation analysis were effected with a 38 c.c. Ge(Li) detector coupled with a 4096 channel pulse height analyser. The detector had a high resolution of 2.2 keV (FWHM) for the 1332 keV 6~ 7-ray and a peak-to-compton ratio of 27. The peak analysis was carried out by using the Hewlett Packard 2116 C computer program. In the radiotracer experiments, the extraction percent was checked by means of a 3" • 3" NaI(T1) connected to a TMC 100-channel analyser. 802

J. Radioanal Chem. 37(1977)

M. H. YANG et al.: DETERMINATIONOF TRACE ELEMENTS Extraction procedure

The acidity of the aqueous phase containing the ions of interest was adjusted by HC104 solution. The organic phase containing HD in n-hexane was preequilibrated with an equal volume of 0.3M HC104. Extraction was performed by adding a fixed volume of organic phase into varying volumes of the aqueous phase in polyethylene flasks. After shaking, aliquots of both phases were taken for the measurements of radioactivity. The extraction percent of a specific ion was thus calculated. Preconcentration

;rhe water samples were first filtered through a 0.45 /am millipore filter to remove suspended particulates. 50 ml of 5" 10-2F HD in n-hexane was added to a 500 ml sample of the water to be analyzed and the acidity of the solution was adjusted to an appropriate value with HC104. After shaking, the organic phase was stripped with a total of 5 ml of 5N HNO3 solution. The stripped acid solution was sealed in a quartz ampoule and served to neutron irradiation. The rice samples were digested in a Sj6strand type reflux apparatus 9 by adding a mixture of concentrated HNO3 and 30% H202. After complete ashing, the solution was neutralized and adjusted to a proper acidity with HCIO4. The extraction was then performed exactly as in water samples. The chemical recovery of the procedures was checked by use of radioactive tracers. Neutron irradiation and radioactivity measurement

Each irradiation set including two samples and one mixed standard solution was packed together in polyethylene rabbits. The samples were irradiated in the OpenPool Reactor at the National Tsing Hwa University (THOR). For the determination of short-lived nuclides, the irradiation was performed in the fast shuttle rabbit for 1 min; whereas for the long-lived nuclides, the irradiation was in the vertical irradiation facility for 30 hours. The neutrgn flux in the irradiation position was about 2 " 1 0 1 2 n - c m - 2 . s e c -I. Following irradiation and suitable cooling period, the samples were transferred from the quartz ampoule and dripped through a HAP column to remove the 24Na produced. Nuclides other than 24Na were eluted with 12N HCI and collected in a polyethylene tube for 3,-ray counting. The irradiated mixed standard solution was poured into a 500 ml unirradiated water sample for subsequent HD extraction. The procedures followed exactly the preconcentration steps as shown above. The final solution was counted under the exact geometrical condition as the samples related to the detector. Z Radioanal. Chem. 37 (1977)

803

M. 1t. YANG et al.: D E T E R M I N A T I O N OF T R A C E ELEMENTS Table 1 The extraction of various elements in natural water by HD solution. (Natural water: 500 ml, HD • 10 -2 F in-hexane: 50 ml) Extraction, % Element

HCIO4 0.180 M

0.018 M

Co

32.6

94.4

9 Cu

36.7

95.3

Zn

31.1

93.4

Cd

71.7

96.7

La

98.6

99.1

Sm

97.9

99.6

AI

78.7

98.3

Fe

62.9

95.3

Mn

19.8

91.4

Hg

6.8

3.0

Na

0.4

5.8

Sb

3.4

3.7

As

0

0

Br

1.8

2.1

Remits and discussion Preconcentration

As expected from Eq. (1), higher HD concentration and lower acidity of the aqueous medium would favor the extraction of a specific ion. Previous study has shown that the plot of lg Kd versus lg (I[ § gave straight lines with slopes of - 1 , - 2 and --3 for metal ions M(I), M(II) and M(III), respectively) ~ In accordance with this, selective extraction of ions of various charge will be possible if suitable control of acidity and HI) concentration are made. In the analysis of trace elements, serious interferences may be encountered with the elements like Na, K, Br, etc. existing in the matrices. Attempts were made to eliminate these interfering elements while still keeping high extraction efficiency for most ions. Table 1 shows a typical result on the extraction of various elements from aque-. ous phase with two different acid concentrations. Extraction was performed with 500 ml of underground water and 50 ml of 5 9 10-2F HI) in n-hexane. The result clearly shows that the extractability of the divalent and trivalent metal ions at low804

J. Radioanal Chem. 3711977)

M. H. YANGet al.: DETERMINATIONOF TRACE ELEMENTS

o

o

U

2O

~.b

6b

H20/HO volume ratio

Fig. 1. The effect of the extraction of 'S2Eu on the volume ratio of aqueous phase to organic phase. Organic phase: 8- 10 -2 F liD in n-hexane, 50 ml; aqueous phase: 'S2Eu in 0.1M H(70, (natural water), varying volume

er acidity (O.O18M HC]O4) were nearly quantitatively extracted while Na, anionic and complex ions (Br, CI, As and Sb) were left in the aqueous solution. One exception among the ions tested is that of Hg. The extraordinarily low extraction may be attributed either to the adsorption" or more probably to the formation of complex ions of Hg in natural water as will be discussed later. To be certain that HD can be a suitable extractant applicable to the determination of metal ions existing in aqueous system at a very low concentration, the effect of the efficiency of extraction on the ratio of the volume of aqueous phase to that of organic phase was investigated. Fig. 1 shows a typical result using i S2Eu as a tracer. The extraction equilibrium in any case was found to be attained within 1 min. It was found that for a 50 ml of 8 ' 10-2F HD solution the extractability is not appreciably affected even if the volume of aqueous phase increases to as large as 1 liter. The result shows also that the HD solution can be successively used to extract new batches of water samples in case the exchange capacity is not saturated by the trivalent ions. For a practical purpose, successive extraction of a series of aqueous phase with HD solution can be proposed. Since there exists relatively large amount of Ca, Mg etc. aside from Na in the bulk of biological and aqueous samples, the investigation of their influence on the extraction of trace concentration of ions of interest would also prove of importance. Fig. 2 shows the extraction characteristics of trivalent metal ion MOII) both in the presence and in the absence of the ions of lower valence M(II) and M(I). The M(III) tested in this work includes Sm and Eu, where M(II) and M(I) are those of Cu, Ca, Mg and K. The amounts of M(II) and M(I) added in the aqueous phase were 150 pg/ml and 500 /ag/ml, respectively. As seen in the figure, the extraction of M(III) is first quantitative and then decreases sharply as the concentration of J. Radioanal Chem. 37 (1977)

805

M. H. YANG et al.: DETERMINATION OF TRACE ELEMENTS

s

~

so

uJ

~M (111)

o|

~

9 M (III) .150 ~ug/ml

M (11)

9

500

/~g/

"~.

oc

4010

A

J

:

'

'

~ ' '100 '

' --'-~

' "*'1000

m''

Fig. 2. The extraction characteristics of M(III) both in the presence and in the absence of M(II) and M(1). Organic phase: 8 ' 10-2F HD in n-hexane, 10 ml; aqueous phase: 1.8. 10 -2 M HCIO, (natural water), 100 ml. Metal ions tested: M01D = S m and Eu, M(II) = Ca and Mg, M(I) = K M(III) increases to a critical value. The extraction behaviour o f M(III) is shown to be nearly identical in the case o f the presence o f M(II) and M(I) as well as in their absence. The fact that the extraction o f M(III) is not appreciable influenced by the existence o f a large amount o f ions o f lower valence in the aqueous phase may ensure quantitative determination o f traces o f trivalent ions in aqueous medium. The HD extract was finally bakc-extracted with a minimal amount of HNO3 for neutron irradiation. The condition for back-extraction can be easily noted from Table 2 that HNOa o f 5M should prove satisfactory. The chemical recovery and reproducibility o f the procedures used in this preconcentration was checked by use o f the radioactive tracers and the results are presented in Table 3. The extraction was performed in 500 ml o f lake water 9 10-2M HCIO4) with 50 ml o f 8" 1 0 - 2 F HD in n-hexane and was back-extracted with 5 M HNOa. All the 10 elements examined are found almost quantitatively recov-

(1

Table 2 The effect of HNO s concentration on the back-extraction of ions from HD extract Back-extraction, %

HNO a , M

F.u / 90.2 | 97.7 98.3 97.7 48.6

806

57.8 90.3 96.1

17.1 71.6 95.0 96.! 96.6 J. Radioanas Chem. 37 (1977)

M. H. YANG et al.: DETERMINATIONOF TRACE ELEMENTS Table 3 Chemical recovery and reproducibility of the preconcentration procedure Element t Recovery,% ] Reproducibility,% .

Sml Eu Cr Cd Zn Cu Fe Co La Sc

i

96.6 95.8 94.6 93.8 91.4 93.5 90.8 94.6 97.8 96.8

i :

1.6 2.1 4.2 5.0 3.8 4.2 4.8 3.0 1.4 3.5

ered (90.8-96.8%). The reproducibility which is given by the relative deviation in percentage with respect to the arithmetic means of three measurements for all elements investigated falls within 5% fluctuation. In the trace determination using HD as a preconcentration agent, special care must be paid to the problem ensuring complete exchange of trace species with the exchanger. Since natural water is an complex system, the trace elements may exist in many forms, e.g. simple ions, complex anions and undissociate compounds. 1-2 Adjustment of the condition by acidification may be necessary to obtain exchange and give reliable results. The recoveries of Zn and Fe in Table 3 were found increased from 91.4% and 90.8% to 95.6% and 96.2%, respectively, if the water sample was treated by acidification (HC104) and evaporation to dryness prior to the HD extraction. The remarkably low percentage of extraction of Hg as was shown in Table 1 may be partly attributed to various complex species in the water medium. Neutron activation analysis

The preconcentration technique developed in this study was employed to determine trace elements in underground water and rice samples. Typical examples of the "r-ray spectra of underground water are shown in Figs 3 and 4. The spectra a and b shown in Fig. 3 are of long-lived nuclides in the sample and spectrum c is that of the blank, which reveales only trace concentration of Sb and Cu. Fig. 4 shows the spectrum of short-lived nuclides. J. Radioanal. Chem. 37 (1977}

807

M. H. YANG et al.: DETERMINATION OF TRACE ELEMENTS A

o)

0

~

o

t-En

OtO

0

L •

_

_

b) 0

8 F~

i 3

G

E ~

0

o ~C

-----/--~..

0

~D

_.t_~L,._ .p.-

0 LJ

0

t

c)

~o

5~

_L 0

L

5o0

I

1000

15()0 Energy , key

Fig. 3. The gamma-ray spectra of the underground water and blank samples (long-lived nuclides). a -- Irradiation: 30 hrs, cooling: 3 days, counting: 50 mins (,sample); b - irradiaton: 30 hrs, cooling: 6 days, counting: 150 mins (,sample); c - irradiation: 30 hrs, cooling: 3 days, counting: 50 rains (blank)

Special care has been paid for the determination of the following nuclides. Since the 844 keV 7-ray from 27Mg and the 846.9 keV 7-ray from 561~dn were superimposed each other, the 1014 keV and 1810.7 keV 7-rays were used for the determination of 27Mg and S6Mn, respectively. For 64Cu determination, the 511 keV annihilation peak was used after subtracting the contribution of 6SZn by referring to the 1115 keV photopeak. 12 For l~sCd determination in rice sample, the 336.6 keV 3'-ray from tile daughter 11Smln was used instead of the 528 keV 7-ray from 11sCd in order to raise the sensitivity by an order of magnitudeJ 3

808

J. Radioanal. Chem. 37 (1977)

M. H. YANG et al.: DETERMINATIONOF TRACE ELEMENTS

%

Q co t.O t_)

C~ t O O ~ t ~ tD r~ t~ aD ~12~c

5o0

~oo

i 1500~

~.o t'3t D

~

O --

t/)

9

tD -

r O

c-4

a~

/

2000

~ ~

2500

/

3000 Energg ~keV

Fig. 4. The gamma-ray spectrum of the underground water (short-lived nuclides). Irradiation: 1 min, cooling: 2.5 mins, counting: 10 mins

Table 4 gives the results of 6 independent analysis of underground water. 13 elements including short- and long-lived nuclides were successfully determined. Reproducibilities for all elements except Fe investigated are within 15%. The analytical results for a rice sample are presented in Table 5. The replicate analysis are shown to give good reproducibility. The data of this study agree with the literature data reasonably well. 14a s

Conclusions The levels of the minor constituents in environmental and ecological samples are sometimes so low that concentration prior to measurement is absolutely essential. In activation method, preconcentration involving enrichment and separation are found necessary before irradiation, in order to avoid high level of high-activity isotopes such as 24Na and 82Br from natural substances. In this study HD has been found as an effective extractant to enrieh heavy metal ions as well as to separate interfering elements such as Na, Br, etc. existing in the natural water samples. The extraction of M(III) is not influenced even in the presence of a large amount of M(II) and M(I) in case the exchange capacity of HD is not saturated by M(III). To ensure complete extraction of trace metals, a suitable treatment of the water samples transferring the metals to the specific exchangeable forms prior to the HD extraction was found to be of importance. Quantitative extraction can be virtually maintained even if the ratio of the volume of aqueous phase to that of HD is as high as 25, thus permitting the enrichment of the trace minor constituents from a large volume of sample water for analysis. J. Radioanal. Chem. 37(1977)

809

M, H. Y A N G et al.: DETERMINATION OF TRACE ELEMENTS

9

O

O

r~

eq r

O r

o

o

o.

O

t~

o~

~1"

,.-4

r162

O O

O O

O O

cS

d

cA

d

d

cS

,-4

g~4

" ,.~ 0 0

O

O

O

O

O

~

O

O

O

O

O

O

O

O

O

O t~

<

.<

O

O

O

O

O

~

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

O

~

O

O

O

O

O

O

6 "U

~.~ 810

J. Radioanal. Chem. 37 (1977)

M. H. YANG et al.: DETERMINATION OF TRACE ELEMENTS The present preconcentration technique has been successfully applied to the determination of trace elements in natural water and biological materials. Both shortlived nuclides (Mg, A1, Ca, Mn) and long-lived nuclides (Sm, Eu, Zn, La, Cr, Sc, Fe, Co) in the ppb level concentrations were determined.

The authors wish to express their hearty thanks to Mr. G. KAISER and Dr. A. WATSON of Max-Planck-lnstitut Ftir Metallforschung, Laboratorium ffir Reinststoffe, Schw~bisch Gmiind, West Germany, for a critical reading of the manuscript. Thanks are also due to Mrs. I. SCHON for her kind assistance in preparing the manuscript.

References 1. J. M. ROTTSCHAFER, R. J. BOCZKOWSKI, H. B. MARK, Jr., Talanta, 19 (1972) 163. 2. R.A. CHALMERS, Enrichment Methods in Trace Analysis in Analytical Chemistry-3 P. Gh. ZUGR~VESCU, (Ed,) London, Butterworths, 1972, p. 569. 3. J.G. MONTALVO, Jr., D. P. THIBODEAUX, E. KLEIN, Anal. Chim. Acta, 52 (1970) i60. 4. T. FUJINAGA, Y. KUSAKA, M. KOYAMA, H. TSUJI, T. MITSUJI, S. IMAI, J. OKUDA, T. TAKAMATSU, T. OZAKI, J. Radioanal. Chem., 13 (1973) 301. 5. S. KAUFMAN, C. R. SINGLETERRY, J. Colloid Sci., 10 (1955) 139. 6. J. M. WHITE, P. TANG, N.C. LI, J. Inorg. Nucl. Chem., 14 (1960) 225. 7. S. M. WANG, N. C. LI, J. Inorg. Nuci. Chem., 27 (1965) 2093. 8. S. M. WANG, N. C. LI, J. Inorg. Nucl. Chem., 28 (1966) 1091. 9. B. SJOSTRAND, Anal. Chem., 36 (1964) 815. 10. S. M. WANG, M. N. CHEN, C. L. TSENG, P.S. WENG, Radioisotopes (Tokyo), 22 (1973) 492. 11. J.M. LO, C.M. WAI, Anal. Chem., 47 (1975) 1869. 12. H. AL-SHAHRISTANI, M. J. AL-ATYIA, J. Radioanal. Chem., 14 (1973) 401. 13. G. H. MORRISON, W. M. POTTER, Anal. Chem., 44 (1972) 839. 14. S. NAGATSUKA, Y. TANIZAKI, Y. OKANO, Radioisotopes (Tokyo) 23 (1974) 693. 15. S.J. YEH, P.Y. CHEN, C.N. KE, S.T. HSU, S. TANAKA, Anal. Chim. Acta, (in the press).

J. Radioanal Chem. 37(1977}

811