‘::I
For y > y 8 and finite k2 the function +(y) cannot conveniently be expressed analytically, and therefore is expressed numerically in tabular form. These numerical values of +(y) were obtained by exactly the procedures described previously ( I ) , following minor modification of the computer programs to replace the potential scans by potential steps.
0.4
0.2-
RESULTS AND DISCUSSION
Figure 1 is a plot of the function +(y) for y , equals 2 [for comparison, Figure 1 includes +(y) for cases in which there is a first order succeeding reaction (2), and no succeeding reaction]. The fact that the current function in Figure 1 for y < y. is independent of kz, whereas for y > y , +(y) is a function of kz (or more generally, y J , permits a number of possible correlations between current and the rate constant. For example, from data for several values of y , a family of working curves directly analogous with those of Schwarz and Shain for the first order reaction could be constructed (2). Any given correlation is essentially arbitrary, however, and might not be preferable in every situation. Thus, the most satisfactory way to present data for the present case is to give values of +(y) for reasonable values of y,. These values of y , are dictated by the fact that to observe optimum influence of the chemical reaction, T should approximately equal the half-life of the reaction, so that values of y, near unity are most useful. Thus, values of +(y) for y > y a are presented in Table I for values of y , in the range 0.5 < y 8 < 3. From these data, curves like Figure 1 can be constructed,
#Y)
0.0-
-0.2-0.4-0.6-
I
0
I
I
I
I
2
3
I I 4
Y ’
Figure 1. Variation of the current function, $,(y), for: ( A ) , first-order succeeding reaction, y = klt: (B), succeeding dimerization, y = kzCo*t: (C) no coupled chemical reaction, y=t and correlations with experimental data appropriate for a given situation can be developed easily. RECEIVED for review November 6, 1968. Accepted February 24, 1968. Work supported by the National Science Foundation and United States Army Research Office-Durham.
Rapid Group Radiochemical Separations for Activation Analysis of Steels Barbara A. Thompson and,Philip D. LaFleur Analytical Chemistry Division, National Bureau of Standards, Washington, D. C. 20234
AMONGthe Standard Reference Materials (SRM’s) certified by the National Bureau of Standards, there are about 100 different types of steel and cast iron ( I ) . Concentrations of 10-20 minor constituents are commonly reported and each must be determined by at least two independent analytical methods before certification. Of the elements whose concentrations are usually certified, Mn, Cu, Ni, Cr, V, Mo, W, Co, Ti, and As could readily be determined by activation analysis if sensitivity were the only consideration. These elements are normally present in steels at levels of a few tenths or hundredths of a per cent which is, for most of them, one or more orders of magnitude higher than is required for detection by neutron activation analysis. In fact, Albert (2) and Malvano and Grosso ( 3 ) have reported the determination of most of these elements in pure iron at the ppm level. (1) “Catalog and Price List of Standard Materials Issued by the National Bureau of Standards,” National Bureau of Standards Miscellaneous Publication 260, U. S . Government Printing Office,Washington, D. C., 1968, p 3-11. (2) P. Albert, Proceedings, 1961 International Conference on Modern Trends in Activation Analysis, College Station, Texas, 1961, p 86. (3) R. Malvano and P. Grosso, Anal. Chim. Acta, 34, 253 (1966). 852
ANALYTICAL CHEMISTRY
The work of Albert and his group utilized extensive radiochemical separations while that of Malvano and Grosso was largely nondestructive. Both of these groups were concerned with the analysis of pure iron; however, most of the steels certified by NBS contain between 0.5 and 2 % manganese and 0.1% or more of copper. The radiation from these constituents completely masks most of the short-lived activities and presents a serious interference for those of intermediate half-life. Additional interferences arise from the iron matrix and from chromium in some of the high chromium steels. Some chemistry is thus necessary in order to obtain the desired information with the high precision and accuracy required for these analytical standards. The detailed radiochemical separation procedures used by Albert have the potential for high precision and accuracy, but are very time consuming and, thus, very expensive. By separating the elements to be determined into groups instead of separating each element individually, and utilizing the high resolution of semiconductor detectors where appropriate, the advantages of nondestructive analysis and radiochemical separations can be combined. Most of the group separation procedures which have been
reported in the literature are based on ion exchange (4, 5 ) and thus are not suited to mixtures containing 100 mg of Fe. It is difficult to remove the Fe without a t least partially removing some of the other elements of interest and so an alternate path was followed in this work. The balance of this paper describes some rapid solvent extraction procedures which have been developed and applied to the analysis of several SRM steels. EXPERIMENTAL
The chemical separation steps were chosen using the criteria that they should be rapid and quantitative, and should result in the minimum number of fractions necessary for precise and accurate determination of the elements of interest. The elements for which the separation scheme was designed were W, Mo, Cu, Cr, As, and Co. Within this group M o must be separated from the Fe matrix to eliminate interference from the 0.143 MeV jgFe peak with the 0.140 MeV 99mTc peak used for determination of Mo. Arsenic must be separated from antimony so that the 0.559 MeV 7 6 Apeak ~ can be measured without interference from the 0.564 MeV 122Sb peak. Also, when small amounts of Co must be determined, the bulk of the Fe should be removed to facilitate accurate measurement of the W o photopeaks. These objectives can be accomplished by the use of two solvent extraction steps. First, the a-benzoinoxime complexes of W and Mo are extracted into CHC13 and, second, Fe and Sb are removed by extraction of their chloride complexes into isopropyl ether. This procedure leaves Cu, As, Cr, and Co in the aqueous phase. These can readily be determined in the presence of each other, utilizing a Ge(Li) detector to separate 0.511 MeV 64Cufrom 0.559 MeV 7+jAs and the 1.17 and 1.33 MeV peaks of W o from the 1.095 and 1.29 MeV 59Fe peaks of any remaining traces of Fe. Other elements such as Au, Ga, and Sb can also be determined, if present, by examination of the ether fraction. A series of tracer experiments was carried out to determine whether this sequence would give the desired precision and accuracy for the elements of interest. To carry out the experiments, it was first necessary to establish a standardized set of starting conditions. Many of the NBS steel and cast iron samples contain large amounts of carbon and silicone.g., 6f with 2.9% C and 1.85% Si-and are difficult to dissolve completely in HC1-HN03 mixtures. In addition, several of the elements to be separated have more than one stable oxidation state. Standardized starting conditions were provided by using the following dissolution procedure. The steel sample (approximately 100 mg) was warmed with 8 ml 1 :1 HN03-HC104 to which 0.5 mg of W, Mo, Cu, As, Co, Cr, and V carriers had been added. (For high chromium steels HCl was used in place of "Os.) When the metal particles had dissolved, about 0.5 ml of HF was added and the mixture was evaporated to fumes of HC104 to volatilize SiF,. For high carbon steels it was rrecessary to continue fuming for an hour or longer before all the carbon was oxidized. The residue was taken up in 10 ml of 6N HC1 and warmed with about 4 drops of HF to redissolve WOs. The volumes were kept as small as possible throughout with the objective of being able to count the final separated fractions directly without concentration. As a result of this dissolution procedure, Cr was in the +6 oxidation state and As and Sb were in the f 5 states. (4) R. E. Jervis and K. Y.Wong, Proceedings of the 1969 Symposium on Nuclear Activation Techniques in the Life Sciences, Amsterdam, International Atomic Energy Agency, Vienna, 1967, p 117. (5) K. Samsahl, D. Brune, and P. 0. Wester, Swedish Atomic Energy Commission Report AE-124, Stockholm, Sweden,October 1963.
Table I. Results of Tracer Experiments Extraction. 7 Element (tracer) a-Benzoinoxime Isopropyl ether 97.1 3= 2.2a . , . Tungsten (187W) Molybdenum ( ~ ~ M o - ~ ~ ~ T 99.4 c ) i0.3 ...
