Instrumental activation analysis of coal and fly ash with thermal and epithermal neutrons and short-lived nuclides

Analytica Chimica Acta, 87 (1976) 451-462 OBlsevier Scientific Publishing Company, Amsterdam -

INSTRUMENTAL WITH THERMAL NUCLIDES

Printed in The Netherlands

ACTIVATION ANALYSIS OF COAL AND FLY ASH AND EPITHERMAL NEUTRONS AND SHORT-LIVED

E. STEINNES Institutt for Atomenergi.

Isotope Laboratories, Kjeller (Norway)

J. J. ROWE U.S. Geological Survey. Reston.

Virginia 22092

(U.S.A.)

(Received 9th June 1976)

SUMMARY Instrumental neutron activation analysis is applied to the determination of about’25 elements in coals and fly ash by means of nuclides with half-lives of less than 48 h; thermal and epithermal irradiations are used. The results indicate that epithermal activation is preferable For twelve of the elements (Ga, As, Br, Sr, In, Cs, Ba, La, Sm, Ho, Wand U). Data for SRM 1632 (coal) and SRM 1633 (fly ash) compare favorably with the results obtained by other investigators_

The combustion of coal is a major source of air pollution, and in the ever-increasing concern regarding environmental contamination, the need has been recognized for versatile and reliable methods for the determination of trace elements in coals and their combustion products. Instrumental neutron activation analysis with thermal neutrons (i.n.e.a.) is one of the most versatile analytical techniques for trace analysis. A comprehensive report by Block and Dams [l] discussed the possibilities offered by this technique for the analysis of coals, based on thermal neutron irradiation and a coaxial Ge(Li) detector. Other investigators [Z-6] have reported analyses of coal and fly ash by similar procedures_ In many cases, the number of eiements that can be determined with acceptable precision and accuracy may be considerably increased by introducing activation with epithermal neutrons in the reactor by using a cadmium cover to eliminate the thermal neutrons. This technique has found several applications to geological materials [7] and was recently applied to air particulate analysis [S] . No applications of epithermal activation analysis (e.n.a.a.) to coals and ashes have been reported to date. In connection with a large scale investigation of trace elements in coals being conducted by the U.S. Geological Survey, a study was initiated to determine whether e.n.a.a. could provide more information than that obtainable by

452

conventional i.n.a.a. based on thermal neutron activation, and to make comparisons when both techniques were applicable. Neutron activation analysis often has been criticized in comparison with other techniques, as requiring a long time from the preparation of the samples for irradiation to the completion of the final analysis. In the present paper, emphasis has been placed on methods involving a short irradiation followed by measurements either after a few minutes or on the subsequent day. Nuclides with half-lives of up to two days have been considered. Thermal neutron activation Block and Dams El], in their study of trace elements in coals, listed 13

elements whose contents can be determined by means of nuclides with halflives shorter than 5 h. From their information on detection limits and concentration ranges in coals, these elements may be classified as follows: 1. elements that can be readily determined in coals by i.n.a.a.: Al, Mg, Cl, Ca, V, Mn, Br; 2. elements that can be determined in a more limited number of coals by i.n.a.a. with satisfactory results: S, Ti, I, Dy; 3. elements which can only occasionally be detected: Cu (by ‘%u), In. According to our experience, Ba, Sr, and U may be added to group 2, but in these cases, more attractive, longer-lived alternative isotopes exist. If the definition of short-lived activities is extended to 48 h, and a measurement on the day after irradiation is performed, Na, K, and La can be included in group 1, while Cu, Eu and Sm may be added to group 2. This probably represents almost the maximum number of elements that can be determined in coalsxia short-lived nuclides by i.n.a.a. with the present state of the art. This observation seems to be verified by the number of elements for which results obtained in this manner have been reported for the National Bureau of Standards Coal (SRM 1632). The situation for fly ash (SRM 1633) is rather similar, except that the halogen concentrations are close to or below the limit of detection. The absence of ‘*Br contributes to slightly improved conditions for elements measured on the second day, as compared with coals. Epithermal

activation

analysis

Activation with epithermal neutrons may often be advantageous if the nuclide of interest has a high I/a, ratio (resonance activation integral to thermal neutron cross-section) in comparison with nuclides giving rise to major activities in the sample on thermal neutron activation. As in many other geological matrices, the major short-lived activities induced in coal and fly ash by thermal neutron activation will, in most cases, be ‘“Al 56Mn, and “Na. Since the stable nuclides from which these radioisotopes $e formed show low I/o0 ratios compared with a number of nuclides of interest in trace element analysis, the activation of coal and fly ash samples within a cadmium cover could, in many cases, lead to substantial analytical improvement.

453

The benefit to be achieved by epithermal activation in a certain reactor irradiation position can be described most accurately by means of so-called “advantage factors” 191, which can be calculated from the relevant activation cross-sections and the cadmium ratio of 19’Au. In cases where a number of radionuclides may be considered as interfering components, the situation may be adequately described by means of I/o0 ratios. Relevant nuclear data for a number of nuclides, formed by (n, 7) reactions, and having half-lives of less than 48 h are presented in Table 1. It appears, from the Table, that e.n.a.a. offers a considerable advantage for the determination of about 25 elements via short-lived nuclides. The feasibility of the epi-cadmium irradiation technique is not only dependent on nuclear data. It also depends on more practical factors such as the reactor operation schedule, the neutron flux characteristics in the irradiation positions accessible for epi-cadmium work, and the extent of the heat build-up by neutron absorption in the cadmium during the irradiation. The possibility of performing e.n.a.a., especially with short-lived activities which require the use of a pneumatic tube system, may therefore be associated with severe limitations in many reactors. In the reactor used in the present work, “rabbits” containing cadmium boxes could be inserted during reactor operation, but the return could only take place after reactor shut-down_ Including the time necessary for postirradiation sample preparation, about 35 min elapsed before the activity measurements could be started_ The use of nuclides with half-lives shorter than about 20 min was therefore ruled out. Preliminary tests on samples of coal and fly ash indicated that the epithermal irradiation technique might be advantageous with respect to the following nuclides, with half-lives less than 48 h: “Ga, 76A~, ‘*Br, 65mSr, “lmSr, 116mIn,12’1, 134mC~, 139Ba, ‘“OLa, ls3Sm, 166H~, “‘W, and 13yU. Peaks of 233Th, 60mC~, “Br and ‘olM~-‘o’T~ were also observed but were not suitable for the determination because of poor peak-to-background ratio. A search for peaks of “Ge, 90mY, 97Zr, 12’Sn, “lEr and 176mL~was negative. Conditions for the measurement of “‘Na, 3sCl, 41-K, 56Mn, “Cu, 15?-mE~and 16sOy were estimated to be more favourable after thermal neutron activation. EXPERIMENTAL Epithet-ma1 activation

Samples of about 100 mg were wrapped in 30-mm X 30-mm sheets of aluminum foil for irradiation. Standards were prepared by evaporating loo-cl1 aliquots of appropriate solutions on to the aluminum foil, drying at room temperature, and folding. Six samples and one set of standards were wrapped in another aluminum foil and placed inside a l-mm tlnick cadmium box (14 mm internal diameter, 10 mm internal height). The cadmium container was wrapped in aluminum foil and irradiated for 30 min in a polyethylene “rabbit” in the pneumatic tube system of the JEEP II reactor

.

454 TABLE

1

Nuclear data of interest nuclides (t&<48 h)

in the epithermal

activation

analysis

00 Na Mg : K Ca Ti V Mn Ni cu Zn Ga Ge As Br Rb Sr Zy, Nb MO

“Na “‘Mg “*Ai ‘EC1 “‘K ‘%a “Ti 52V *&Mn “SNi *‘Cu 66cu 69mZn 72Ga ‘5Ge TbAs ‘OBr “Br =Rb *7mSr XlIllY 97Zr-97Nb g’mNb lCl.M_*Cl*.l.c

In Sn I cs Ba

I *amIn *25mc& ,**)I 13’IIlCs I 35mga

La Nd Sm

la9Ba 140La li9Nd ’ %rn *srSm

Eli Gd DY Ho Er Yb Lu W Th U

LS’m& ls9Gd ‘=Dy lb6Ho l”Er “‘Yb

I 76m~u 1tl7W “3Th 239~

15.0 h 9.5 min 2.3 min 37.2 min 12.4 h 8.8 min 5.8 min 3.76 min 2.58 h 2.56 h 12.8 h 5.1 min 13.8 h 14.2 h 82 min 26.5 h 17.6 min 35.3 h 17.8 min 2.83 3.1 17.0 6.3 14.6

h h h-74 min min min-14 min

54.0 min 9.7 min 25.0 min 2.9 h 29 h 82.9 min 40.2 h 1.8 h 47.0 h 23 min 9.2 h 18 h 139 min 26.9 h 7.5 h 1.9 h 3.69 h 23.9 h 22.2 min 23.5 min

1101

of coals based on short-lived

z r111

0.528 0.038 0.232 0.43 1.48 1.1 0.179 4.30 13.3 1.50 4.4 2.20 0.07 4.7 0.52 4.4 8.4 3.0 0.12 0.8 0.001 0.05 0.15 0.20 161 0.14 6.2 2.6 0.16 0.35 9.0 2.5 210 5.5 2800 3.5 2600 61.5 6 5.5 18 37 7.4 2.72

0.31 0.030 0.17 0.17 1.28 0.90 5.5 2.7 14 0.9 5.0 2.5 0.24 25 0.9 63 125 50 2.3 4 0.9 5.0 8.0 3.9 2600 9 150 30 24 0.3 11 17 2900 27 2600 80 800 660 35 7 600 420 82 280

0.59

0.79 0.73 0.40 0.86 0.82 37 0.55 1.1 0.6 1.1 1.1 3.4 5.3 1.7 14 15 17 19 5.0 900 100 53 20 16 64 24 12 ,_ 15 0.9 1.2 6.8 14 4.9 0.93 23 0.31 11 5.8 1.3 33 11 11 103

455

(Kjeller, Norway) at a thermal flux of about 1.5 - 1013 n cmcadmium

s-l and a ratio of 19’Au of 2.9. Insertion took place 30 min before reactor

shut-down and the “rabbit” was retained in the reactor for another 20 min before return to the laboratory. After unpacking, the samples were weighed and wrapped in inactive aluminum foil. Activities were measured with a system consisting of two coaxial Ge( Li) detectors (5 9%relative efficiency, 2.5 keV FWHM and 3 % relative efficiency, 3.0 keV FWHM, respectively) with associated electronics, interfaced to a NORD I computer (16 K, 16 bit). Two samples or standards were counted simultaneously for 10 min. The distance between samples and detector was so adjusted that the dead-time did not exceed 15 %. Dead-time corrections were made by means of precision pulse generators, and the peak areas were calculated by the method of Cove11 [12]. In some cases only one channel on each side of the top channel was included in the peak area in order to avoid interferences from adjacent peaks. The samples were re-counted after an interval of about 24 h. The radionuclides, decay times, and -y-energies involved are given in Table 2. Activation

without

cadmium

cover

Samples of about 100 mg were sealed in small envelopes of polyethylene foil. Standards were made by absorbing 50 ~1 of a solution on to filter paper and sealing in the same sort of polyethylene envelopes_ Two samples or standards were irradiated in each “rabbit”. The irradiation time was 5 mm

TABLE

2

Elemental concentration in NBS Coal (SRM instrumental Element

activation Nuclide measured

Ga

‘=Ga

As

16&

Br Sr

82Br ssmSr “mSr

In

I lamIn 3=mcs

CS

I

Ba

ls9Ba lJoLa ’ ‘%m ‘6LH~ *StW “9U

La Sm Ho W U aMean

of 5 separate

analysis

1632) and NBS Fly ash (SRM with epithermal neutrons (p.p.m.)

T-Energy (keV) 630 834 559 618 231 388 417 1099 127 165 1596 69 80 685 74 determinations.

Decay time(h) ‘L 24 24 24 24 % 1 1 1 1 1 1 % 24 24 24 24 ,x 1

NBS

1632

Mea+ 6.1 5.8 6.5 19.5 165 161 0.0169 0.0178 1.36 338 11.4 1.38 0.24 0.79 1.46

’ sr (B) 10.7 7.7 18.5 1.5 12.7 5.6 7.1 7.8 7.4 4.1 4.3 7.2 12.5 21.5 1.4

1633)

NBS

by

1633

Mean” 40.3 40.7 58.0 7.7 1360 1430 0.128 0.118 8.2 2540 81.2 11.4 1.94 4.2 12.7

. sr (%I 5.0 3.0 3.5 19.5 8.1 4.2 6.3 3.4 11.0 2.0 4.0 14.0 6.7 9.5 4.0

456

for coal samples and 1 min for ash samples. After 20 min, measured with a stopwatch, the samples were counted for 10 min as described above. The interval between subsequent irradiations was 15 min to allow sufficient time for concentration print-outs and changing of samples. After 24 h, the samples were recounted, in the same sequence, to avoid time-consuming calcuIations_of decay corrections. The radionuclides, decay times and y-energies used for this part of the investigation are listed in Table 3. .

.

RESULTS

AND DISCUSSION

In order to compare the two methods and obtain a measure of the precision and to some extent the accuracy to be expected on application to coals and ashes, the standard reference materials coal (SRM 1632) and fly ash (SRM 1633) from the U.S. National Sureau of Standards were investigated_ Five replicates of each sample were analyzed and the results are given in Tables 2 and 3 for the epithermal and thermal neutron irradiations, respectively. In Tables 4 and 5, the “best” values from the present work, as indicated in the following discussion, are compared with literature values for SRM 1632 and SRM 1633. The methods also have been TABLE

3

Elemental concentrations in NBS Coal (SRM 1632) and NBS Fly ash (SRM 1633) instrumental activation analysis with thermal neutrons Element

Na, p_p.m. Mg, % Al, % Cl, p.p_m_ K, % Ca, % Ti, % V, p-p-m_ Mn, p-p-m. Cu, p_p.m. Br, p-p-m. Sr, p-p-m. I, p.p_m_ Ba, p-p-m_ La, p-p-m_ Eu, p-p-m_ Dy, p-p-m. U, p-p-m_

by

Nuclide measured

Y-Energy (keV)

Decay time

NBS 1632 Mea@

sr (%)

Mean=

sr (%)

“Na “Mg =Al “‘Cl 42K “Ca “Ti “V 56Mn 6JCu “‘Br s’mSr “‘1 “9Ba la”La ‘52mEu “‘Dy “‘9U

2756 1014 1780 2167 1524 3084 320 1434 1811 511b 617 388 443 165 1596 122 94 74

Id 20 min 20 min 20 min Id 20 min 20 min 20 min 20 min Id 20 min 20 min 20 min 20 min Id Id 20 min 20 min

380 0.17 l-74 844 0.298 0.350 0.089 35.0 41.1 15.0 14 131 2.9 274 10.3 0.299 1.12 1.52

6.5 17.6 2.3 4.4 8.0 8.0 5.6 8.3 8.8 8.0 14.3 17.6 10.3 11.3 10.7 11.0 5.4 7.2

2830 1.78 12.35 -

4.8 11.2 2.0

1.80 4.69 0.70 237 488 115

7.2 3.0 4.3 8.4 2.9 7.0

NBS 1633

-

1200 2580 91 2.44 9.4 13.5

aMean of five separate determinations_ bAfter subtraction of contribution from pair production associated with **Na.

25 6.6 7.7 7.8 5.3 8.9

“‘I’= thermal neutron

activation,

E = epithermal

380 f 25 T 414 f 20 0.17 f 0.03 T 0.20 f 0.06 1.74 i 0.04 T 1.85 f 0.13 844 f 37 !I 890 I 125 0.298 c 0,024 T 0.28 i 0.03 0.350 !: 0.028 T 0.43 !: 0.05 0.089 i 0.005 T 0.104 + 0.011 35.0 r 2.9 T 36 ?: 3 41.1 ” 3.6 2’ 43 f 4 15.0 f 1.2 T 5,8 f 0.4 E 6.5 f 1.2 II” 6.5 f 1.4 19.5 f 0.3 E 19.3 f 1.9 161 f 9 E 161 1: 16 0.0169 ” 0.0012 E 0.20 f 0.12 2.9 k 003 T 1.4 * 0.1 1.36 !: 0.10 E 338 f 14 E 352 i 30 11.4 -+ 0,5 E 10.7 2’1.2 1.7 f 0.2 1.38 f. 0.10 E 0.299 f 0.022 T 0.33 f 0.04 1.12 i: 0.06 T 0.24 ” 0.03 E 0.79 ?: 0.17 E 0.75 + 0.17 1.46 f 0.02 E 1.41 i 0.07

Na, p.p,m. Me, % Al, % Cl, p.p.m. K, % Ca, % Ti, % V, p.p.m. Mn, p.p.m. Cu, p.pm. Ga, p.p.m. As, p.p,m. Br, p.p.m, Sr, p.pem. In, p.p.m. I, p,p.m, Cs, p.p.m. Ba, p.p.m, La, p.p.m. Sm, p.p.m. Eu, p.p,m. Dy, p.p.m. Ho, p.p.m. W, p.p.m. U, p.p,m,

Ondov et al. [3]

Present work preferred values*

neutron

0,74 i 0.3 1.46 j: 0.35

5.15+ 0.3 8.9 + 045 18.2 f 2.3 145* 9 0.22 f 0.02 2.8 .t 0.4 1.8 f 003 366 i 34 11.3 .i 0.4 1.8 * 081 0,41 f 0.03 1.3 1: 0.5

1.26

0.21

1.4 405 lo,5

value.

390 0.248 1.90 1000 0.290 0.44 0,093 402 3 46i 3 18 8.5 5.5 1402 144 0.07

activation. bInformation

? 1.2 f 0.11 t 25 * 0.15 * 0016 + 0.02 k 0.09

1.2 z! 0.1

6.63 1.32 311 7.89 1.66 0.37 1.38

4.61 i 0832 15,2 t 1.4 1.02 +- 0.05

43.7 i 1.8

846 f. 44 0.29 * 0.02

383 f 14

[141

Klein et ai. [ 61

values

1.37 !: 0.08

0.35 i 0.04 314 ?. 20

1.33 i 0.1 0.23 f 0.02

5.25 i 0.37

0.0973 * 0.0050 33.9 f 3.0 47.1 i 4.1

930 f 48

351 !: 30 0.160 2 0,015

Chattopadhyay and Jervis [ 131

of present data with literature

[ 51 Ruth et al.

347 i 32 0.15 !. 0.03 1.76 f 0.31 945 * 35 0,278 i 0.023 0.43 * 0.02 0.084 i 0.017 3207 e 3,4 40.3 f 6.9

Nadkarni

in NBS Coal (SRM 1632). Comparison .-_---

Element

Elemental concentrations

TABLE 4

1.43

2.6 280 11.3 1.6 0.4 1.4

129

46

0.29

410

Millard and Swanson [ 41

1.4 f 0.1

5.9 f 0.6

0.080b 35 1 3 40 f 3 18 f 2

NBS Certified value

2830 x 140 2 1.78 f 0.20 T 12,36 f 0.26 T 1.80 ?: 0,13 I’ 4.69 i 0.14 ‘I 0.70 f 0.03 9’ 237 i 20 T 488 k 14 T 115 :! 8 T 40.3 * 2,0 E 58,0 k 2.0 I!,’ 7.7 s 1.5 E 1430t 60E 0,128 t 0.008 E 8.2 f 0.9 E 2540 ?: 50 E 81,2 t 3.3 E 1104 * 1.6 E 2.44 2 0.19 T 9.4 + 0.5 T X.94 t 0.13 E 4.2 f 0.4 E 12.7 f 0.5 6

Na, p,p.m,

4.6 zt 1.6 12.0 i 0.5

67.6 i 0.9 7.62 ‘, 0.46 1406 f 80

58 +-4 122 4 1700 1300 0.32 f 0.10 8.6 f 1.1 2700 t 200 82 f 2 12.4 f 0.9 2.5 i 0.4

11.3 f 0.3

5.81 f 1.4 2700 i 200 64.1 + 1,6 13,6 t 0.88 2,62 f 0.05 10.9 f O&O

3300 t 200 1.62 f 0.06 14,3 1 1,l 1678 i 0,23 4.21 t 0.09 0,61 :! 0.02 223 t 9.9 464 + 1.4

3200 f 400 108 + 0‘4 12.7 i 0.5 1.61 * 0,16 4.7 2 0.6 0.74 f: 0.03 235 !: 15 496 * 19

[51

[31 -

Nndlcarni

Ondov et nl,

“T= thermal neutron activation, E = epithcrma1neutron ~ctivatian.

Mg, % Al, % K, % Ca, % Ti, % V, pepam. Mn, p.p.m. Cu, p.p.m, Ga, p.p.m, As, p,p.m. Br, p.p.m, Sr, p.p,m, In, p.p.m. Cs, p.p.m, Ba, p,p.m. La, p.p.m, Sm, p.p.m, Bu, p.p.m, DY, p.p.m. Ho, p-p-m. W, p.p.m. U, p.p.m.

Present work, preferred valuen

Element

11.8

2780 82 15 2.86

12.6 1.8 4.34 0,642 240 460 133 49 54 6.0 1301

5 m

11,7

3.1 lo,2

9.4 2490 80

11.6 f 0.2

1380 1390

1373 f 95 0.28 ?: 0.03 0.63 f 0.06 2610 i 210

214 f 8 493 f 7

1.72

3070 I,98

611 6

540

1,77

3400

Millard and NBS Swnnson [ 41 Certified value

60.7 f 2,6

3,92 1: 0.28 0.723 i 0,04 208 t 12 495 ‘; 25

3400 t 300 1.48 f 0.1

Klein ct al. Chnttopadhyay and Jervis [ 131 Z6l

Elemental concentrations in NBS Fly ash (SRM 1633). Comparison of present data with liternture values

TABLE 6

459

applied to a study of the fractionation of trace elements during combustion in coal-fired power plants, the results of which will be reported elsewhere

[I51 f

A separate discussion of each of the elements studied is given in the following paragraphs. Na, Mg, Al, Cl, K, Ca, Ti, V, Mn These elements were determined by thermal neutron irradiation, and the

results agree with previously published values. The precision observed in this work is similar to that reported by other investigators_ 1.n.a.a. seems to yield rapid, reliable results for Na, Al, V, and Mn in coal and fly ash, and for Cl in coal. For K, Ca, and Ti, the situation appears to be less satisfactory and for Mg, relatively poor results can be expected_ Chlorine was not detected in the fly ash samples. The only point where the present procedure deviates from most previous work with regard to the above elements is that Al is determined along with 12 other elements in a 10 min count starting 20 min after the irradiation so as to avoid making two different counts for this group of elements. The activity of *‘Al in the first few minutes of the counting period is still sufficiently high to allow a precise determination (cf. Tables 2 and 3), but it is not so high as to cause a major contribution to the dead-time. The only element in this group that might be more advantageously determined by e.n.a.a. is titanium. Because of the short half-life of 51Ti (5.8 min), e.n.a.a. was not attempted for this element. Cu, Ga, As, Br, Sr, In, I, Cs, Ba Copper. Copper as determined by i.n.a.a. has not been reported previously

for these materials. The result obtained in this work for SRM 1632, after proper correction for the interferences from “Na to the 511-keV peak of 64Cu, is in agreement with the certified value within error limits. Gakm. The determination of gallium in SRM 1632 and SRM 1633 by i.n.a.a. has not been previously reported. The determination of gallium can readily be accomplished by e.n.a.a. because of the favorable I/a,, of 7’Ga relative to that of 23Na. Our value of 5.8 _+0.4 p.p.m_ for SRM 1632 is in fair agreement with the 5.15 + 0.3 p.p.m. value by Ruth et al. [14] obtained by a radiochemical neutron activation procedure. Arsenic The e.n.a.a. determination of arsenic in both the coal and fly ash sample was accomplished after only a 24-h decay. In order to obtain similar results by thermal neutron activation, it would have been necessary to wait for an additional 2-3 days before making the measurements. Bromine. Bromine was determined easily in both samples via “Br after only 24 h decay by means of e.n.a.a. At this time, the determination by thermal neutron activation was more difficult, especially in the case of the fly ash sample. An attempt to determine bromine by means of “Br after thermal neutron activation was not successful. Strontium. Two different isotopes (85mSr and ““Sr) could be used for the

460

determination by e.n.a.a. The results obtained by means of x7mSrappeared to show higher precision for both the materials. The values for SRM 1633 are in good agreement with the certified value. By thermal neutron activation, very poor results were obtained with ti7mSr,and ssmSr could not be detected_ Indium. Indium is especially suitable for analysis by e.n.a.a. The advantage to be obtained relative to 56Mn, the major interfering nuclide, is about a factor of 15. This makes it possible to determine indium in coals with a precision of about 10 % and in ashes with a precision of about 5 %, without the necessity for a chemical separation. Both the 417- and 1099-keV peaks are applicable. Attempts to use the 1293-keV peak gave systematically high results which may be associated with the presence of trapped 41Ar which has a coinciding y-ray. According to Block and Dams [l] , the concentration level of indium in coals is of the order of 0.01-0.03 p-p-m. Application of e.n.a.a. to U.S. coals gave values of similar order here. Several investigators have reported values around 0.2 p.p.m. for the indium content of SRM 1632. We believe that those results may be erroneous, and may be due to spectral interferences, contamination, or other sources of systematic errors. Iodine. The iodine content of SRM 1632, determined by thermal neutron activation, was in agreement with the value reported by Ruth et al. [14]. Theoretically, the sensitivity for iodine determination in coals and ashes should be about a factor of 20 improved by using e.n.a.a. In attempts to prove this, difficulties were encountered from volatilization of “‘I from the standards during irradiation_ This experiment was therefore discontinued. Cesium. The determination of cesium by neutron activation analysis is usually based on 2-3-y 134Cs,which shows a high sensitivity by i.n.a.a. and even better for e.n.a.a. In the present work, the cesium determination by e.n.a.a. was shown to be feasible by means of the short-lived isotope ‘34mC~. The present results compare well with the values from the joint study by Ondov et al. [3] _ Barium. Although the determination of barium via 13’Ba in these samples by i.n.a.a. is possible, the reproducibility was found to be significantly improved by e.n.a.a., as was expected from the nuclear properties of 13’Ba. La, Sm, Eu, Dy, Ho, W, U Lanthanum. The results from this study agree with the values in the literature. E.n.a.a. appears to show better precision than i.n.a.a. when the measurements are carried out after one day, primarily because of a higher ““La/‘4Na ratio. Samarium. The present values, based on measurement of the 69-keV y-ray after epithermal irradiations, are slightly lower than most of the literature values. This could be related to spectral interference from 239Np when the 103-keV y-ray of ls3Sm was measured in some of the other work. This interference could be significant because the uranium content of the samples is fairly high.

461

Era-opium and dysprosium_ These elements are determined best by thermal neutron activation. The present data agree with most of the literature values, except for dysprosium in SRM 1632, where our value is slightly lower. Holmium.

The present data for holmium

seem to be the first reported

for

SRM 1632 and SRM 1633. The main reason for this is primarily the advantage obtained by using e.n.a.a. Tungsten. Tungsten can be determined in many coals and ashes by i.n.a.a. and several values for SRM 1632 and SRM 1633 have been reported. Also in

this case, e.n.a.a. is to be preferred because the I/a, ratio of lg6W is almost 20 times that of ‘%a: Uranium. This-clement is one of the very best cases for e.n.a.a. The “advantage factor” of 23gU relative to the major interfering nuclide 56Mn is calculated to be 24 under the present conditions. Even though uranium could be determined in SRM 1632 and SRM 1633 by i.n.a.a. with this nuclide, the results obtained by e.n.a.a. show significantly greater precision. Shielding effects The possibility of shielding effects should never be overlooked in activation analysis. Possible errors from such effects in e.n.a.a. have been discussed elsewhere [7] in the case of silicate rocks. Since the elemental composition of coal ash and silicate rocks is somewhat similar, the shielding effects should also be insignificant in this case and even more so in the case of the coals because of the low neutron capture cross-section of carbon. However, carbon is a good neutron moderator and one might expect some moderation of epithermal neutrons in the samples. The possible effects of such neutron moderation seem to be negligible, however, since results from the two different techniques appear to be in good agreement both in the case of a nuclide with a very high I/a, ratio (“3sU) and a nuclide with a fairly low value for this ratio ( 13’La). In case of larger sample weights, neutron moderation effects may prove to be significant. The thermal neutrons thus produced would be likely to be absorbed by the cadmium, and the net result might be a lower exposure of epithermal neutrons to the samples than to the standards. In any case, errors of this kind could be avoided by using a coal as a standard. As demonstzated in this work and other recent investigations, many shortlived nuclides may be advantageously used in the multi-element analysis of coal and its combustion products by i.n.a.a. The introduction of e.n.a.a. as a supplementary technique extends the number of elements that can be determined by the purely instrumental approach and facilitates an improvement in the results obtained for a number of the elements compared with those obtained by i.n.a.a. with thermal neutrons.

462 REFERENCES 1 C. Block and R. Dams, Anal. Chim. Acta, 68 (1973) 11. 2 R. R. Ruth, H J. Gluskoter and N. F. Shimp, Occurrence and Distribution of Potentially Volatile Trace Elements in Coals, Environmental Geology Notes No. 72, Illinois State Geological Survey, Urbana, 1974. 3 J. M. Ondov, W. H. Zolier, I. Omez, N. K. Aras, G. E. Gordon, L. A Ranciteili, K. H. Abel, R. H_ Filby, K. R. Shah and R. C. Ragaini, Anal. Chem., 47 (1975) 1102. 4 H. T. Millard and V. E. Swanson, Trans. Am. Nucl. Sot., 21 (1975) 108. 5 R. A. Nadkarni, Radiochem. Radioanal. Lett., 21(1975) 161. 6 D. H. Klein, A. W. Andren, J. A. Carter, J. F. Emery, C. Feldman, W. Fulkerson, W. S. Lyon, J. C. Ogle, Y. Talmi, R. I. van Hook and N. Bolton, Environ. Sci. Technol., 9 (1975) 973. 7 E. Steinnes, in A. 0. Brunfelt and E. Steinnes (Eds.), Activation Analysis in Geochemistry and Cosmochemistry, Universitetsforlaget, Oslo 1971, p_ 113. 8 M. Janssens, B. Desmat, R. Dams and J. Hoste, J. Radioanal. Chem., 26 (1975) 305. 9 D. Brune and K. Jirlow, Nukleonik, 6 (1964) 272. 10 R. Sher, in Handbook on Nuclear Activation Cross-Sections, pp_ l-13, Tech. Rept. Series No. 156, International Atomic Agency, Vienna, 1974. 11 H. Albinsson, in Handbook on Nuclear Activation Cross-Sections, pp_ 15-86, Tech. Rept. Series No. 156, International Atomic Agency, Vienna, 1974. 12 D. F. Covell, Anal. Chem., 31 (1959) 1785. 13 A. Chattopadhyay and R. E. Jervis, Anal. Chem., 46 (1974) 1630. 14 R. R. Ruth, R. A Cahill, J. K. Frost, L. R. Camp and H. J. Gluskoter, Trans. Am. Nucl. Sot., 21 (1975) 107. 15 J. J. Rowe, E. Steinnes and P. Zubovic, to be published_