Methodology for high-flux absolute multielement neutron activation analysis — environmental baselines by analysis of tree rings

Multielement determinations by high-flux absolute n.a.a. require careful methodology to avoid systematic error. The ORNL high-flux facility with a thermal flux of 5 ·10¹⁴ cm^-2 s^-1 and a thermal-to-resonance flux ratio of

T004. Analysis of Knoxville tree-core samples^a

Element	Increment no.20 (1832–1838)	Increment no.8 (1932–1838)	Increment no.3 (1858–1863)	Increment no.1 (1868–1872)
					Core Number	Core Number	Core Number	Core Number
	1	3	1	3	1	3	1	3
Na	12.9(1.1)	2.67(3.3)	33.8(1.0)	11.2(1.6)	15.6(1.5)	31.4(1.2)	22.4(0.93)	9.9(1.8)
Mg	c	c	160(12)b	80(27)b	100(21)b	100(30)b	c	c
Al	15.2(5.4)	5.6(6.1)	12.5(3.8)	11.0(4.4)	7.8(6.6)	8.4(7.4)	6.5(14)	6.6(6.2)
Cl	15.4(9.1)	ND	66(5.4)	ND	ND	55(6.1)	26(5.5)	22(5.1)
K	205(2.6)	223(3.3)	493(1.8)b	457(1.9)b	478(1.9)b	439(2.1)b	449(1.8)	570(1.7)
Ca	1600(11)	ND	ND	ND	ND	ND	ND	ND
Cr	ND	ND	NDb	NDb	0.3(>30)b	0.6(>30)b	0.6(>30)b	ND
Mn	109(0.30)	107(0.23)	64.1(042)	108(0.32)	46.7(051)	66.8(0.44)	38.9(0.39)	45.0(0.34)
Fe	ND	ND	ND	ND	ND	19(19)	25(18)	ND
Co	ND	0.02(>30)	ND	ND	ND	ND	0.30(>30)	ND
Zn	7.2(7)	7.4(4.2)	13.0(2.4)	9.2(3.3)	5.6(4.0)	4.5(4.7)	ND	2.4(8.8)
Br	0.10(20)	ND	0.07(12)	ND	0.024(13)	ND	0.04(>30)	ND
Rb	ND	ND	0.9(13)	0.8(18)	0.8(15)	0.7(17)	ND	1.1(18)
Ag	ND	ND	1.06(3.0)	15.3(0.71)	0.15(15)	0.69(4.8)	0.11(28)	ND
Ba	11(7.3)	10(13)	3.2(16)	ND	3.0(15)	2.9(18)	ND	3(32)
W	ND	ND	ND	ND	ND	ND	0.06(29)	ND

a

Results in p.p.m followed by (per cent counting statistical error).

b

Computer forced result.Also V,Se,Sr,Sb,Cs and au (forced) were not detected.

c

Element possibly present but missed by peak-finding routine.7 35–45, was used in developing an instrumental absolute multielement method. The detector was calibrated for absolute counting with two independent sets of radioactivity standards for four detector- -source distances; the absolute activities of the standards were reproducible within accuracies of 9%. Five sources of systematic error were investigated: (a) correction for counting of cylindrical sources for 26 γ-ray energies reached 14–17% for photon energies below 500 keV; (b) flux variation during bombardment and within the irradiation capsule volume was not significant; (c) samples were sufficiently stable during high-flux bombardment; (d) multi-element impurities in accessory materials (polyethylene and “Nucleopore” filters) were not significant; (e) correction for sample activation during rabbit transfer was necessary for short bombardments, e.g., 8.6 % for 6 s and 19.6 % for 4 s. This methodology resulted in accuracies of 10–15 % for most elements, as determined by analysis of N.B.S. orchard leaves and coal and of Bowen's kale standards. The method was applied to a preliminary chronological study of environmental baselines and contamination levels, based on tree ring samples, covering a period of 100 years.