Nuclear Chemistry Division, Pakistan Institute of Nuclear Science and Technology, p.o. N/lore, ... manufacturing of aircraft engine turbine blades and chemical.
Journal of Radioanalytical and Nuclear Chemistry, Vol. 242, No. 2 (1999) 259-263
Determination of trace impurities in nickel-based alloy using neutron activation analysis J. H. Zaidi, S. Waheed, S. Ahmed Nuclear Chemistry Division, Pakistan Institute of Nuclear Science and Technology, p.o. N/lore, Islamabad, Pakistan (Received September 28, 1998)
A radiochemical neutron activation analysis procedure has been applied to investigate 40 major, minor, and trace impurities in nickel-based alloy. The extensive use of these alloys in the electronic industry, telecommunications, manufacturing of aircraft engine turbine blades and chemical equipments desires for their precise characterization. The concentration of nickel in the nickel-based alloy was found to be 56.8%, whereas Fe, Cr, Ca, Mg, Ce, Mn, Na and V were the major components o f the alloy, which constituted to more than 26%. The rest of the elements was present in minor or trace levels. Most of the rare earth elements except Ce were also present in trace amounts. Neutron activation analysis technique was preferably used because of its good sensitivity and multielement determination capabilities for the characterization of high purity materials. The comparison of RNAA and 1NAA indicated improvement in the detection limits utilizing radiochemical separation procedures developed in the present work.
Introduction In the manufacture of nickel-based alloys for hightemperature application, such as aircraft engine turbine blades, close control is required of certain low boilingpoint trace elements. These elements can have a deleterious effect on the mechanical properties of the alloy; therefore, lower permissible levels are continually being demanded. Production techniques such as vacuum refining with close control of raw materials are used to manufacture these alloys. The effect of such techniques is measured by the determination of the impurities. 1 The trace impurities in high purity materials can be determined using different analytical techniques such as atomic absorption spectrometry (AAS), inductively coupled plasma mass spectrometry (ICPMS), positron induced X-ray emission (PIXE), photon activation analysis (PAA), charged particle induced activation analysis (CPAA) etc. Neutron activation analysis (NAA) has long been used for such analyses. However, a careful consideration of the sources of uncertainty involved using this technique reveals a significant list of potential contributors. 2,3 Many efforts have been made to develop standardization procedures that minimize some of the contributions to the uncertainty to achieve high accuracy. Some investigators have proposed and refined the variant of the "intemal comparator method" by which the sample itself is used as a material in the preparation of the comparative standa r d. 4-8 The method based on kofactors (k0 is a combination of some nuclear parameters that characterize the activated element) has been developed by SIMONITS et al., 9,1~ and permits a more accurate determination of the elemental concentrations by using only one or two comparators since the tabulated ko-values are directly determined experimentally.
0236-5 731/99/USD 17.00 9 1999 Akaddmiai Kiad6, Budapest All rights reserved
Comprehensive descriptions of such standardization techniques as applied in neutron activation analysis have been given by DE CORTE.I 1 The use of comparators is avoided when applying the semi-absolute method 12 or the mathematical standardization procedure advanced by CINCU,13,14 which is mainly applicable to materials produced by industrial or special technologies. Under such cicumstances,an absolute calibration of the spectrometric setup is required.The relative method, on the other hand, requires a standard to get reliable results; such a reference (certified reference material containing specific concentrations of several elements) should be identical to the sample with respect to gross composition, shape, volume and other physical characteristics. To improve the sensitivity of the technique radiochemical separation procedures are employed) 5,16 The relative RNAA based technique has been extensively used in our laboratory for the analysis of high purity materials, ores, rocks and biological materials. 17-19 In the present work a radiochemical separation procedure has been developed and applied to the determination of trace impurities in high purity nickel-based alloy. Experimental
Samplepreparation The samples of high purity nickel-alloy foil of 0.25 mm thickness were taken. To avoid contamination special care has been taken for the cleaning of the tools used for sampling. Sample containers and all the samples were washed with trichloroethylene, acetone and deionized water, sequentially. Then the samples were etched with 0.1M HCI to remove any surface
Elsevier Science B. K, Amsterdam Akaddmiai Kiad6, Budapest
J. H. ZAIDIet al.: DETERMINATIONOFTRACEIMPURITIESINNICKEL-BASEDALLOY
contaminations. Finally the samples were rinsed with deionized water, ethylalcohol and acetone. The reagents used were of electronic grade. All the procedures upto the final sealing of the samples was carried out in a clean bench facility to avoid any possible external contamination.
Standard preparation The standards were prepared from the stock solutions containing 1 mg m1-1 of the respective elements under investigation. These solutions were diluted accordingly to give a range of standards for each element from 5 ppm to few ppb. The solutions were dried on ashless filter paper and sealed in polyethylene/quartz capsules for irradiations.Blank filter papers were also irradiated to subtract their contributions if any. Standards of appropriate concentrations for all the elements under investigation were used, i.e., according to the level of respective impurities in the samples.
Neutron irradiation Samples, each weighing about 200 mg, were taken in triplicate and heatsealed in pre-cleand polyethylene and quartz ampoules for short and long irradiations, respectively. The quartz ampoules were then placed in NRX type irradiation capsules and cold-welded. For long irradiations a 10 MW swimming pool type research reactor (PARR-I) was used whereas for short irradiations a 27 kW tank-in pool type research reactor (PARR-II) was used. The thermal neutron flux densities at the irradiation sites of the two reactors were of the order o f 5-1013 n-cm-2s -1 and 1012 n ' c m - 2 s -1, respectively. Thermal neutron flux monitors, e.g., Au, Co, and A1 foils, were inserted between the samples and the standards to monitor the fluctuations in the thermal neutron flux gradient; which were found to be insignificant. The samples along with appropriate amounts of standards were irradiated for 2 minutes to 24 hours according to the requirement. The irradiated samples and standards were transferred to the preweighed polyethylene vials and re-weighed to determine the exact weight.
Radiochemical procedures Radiochemical separations are required for the determination of shortlived activation products, such as 28AI, 116In and 27Mg, to avoid masking due to high matrix activity. A radiochemical separation procedure has been developed as detailed below. Irradiated nickel-alloy samples were dissolved in 2 ml of aqua regia, evaporated almost to dryness and redissolved in 1 ml of 10M HC1. The mixture was loaded on a pretreated 1 cm 0• cm long Dowex 1X8 resin
260
column (C1 form, 100-200 mesh). The major constituents of alloy, i.e., nickel along with REE, A1, Br, Cs, In, K, Mg, Mn, Na, Rb and Sc, were eluted with 10M H C 1 . Nickel was then precipitated with dimethylglyoxime (DMG) (solution in ethyl alcohol, DMG was almost seven times the weight of Ni). The precipitate was filtered and kept at 110-120 ~ for 2 hours and mounted for y-spectrometry. The filtrate was diluted with 5 ml of deionized water and pH was adjusted to 3 with 25% ammonia solution. Then 2 ml saturated solution of oxalic acid was added to precipitate rare earths and scandium and was subjected to y-spectrometry after filtration under vacuum. The filtrate containing A1, Br, Cs, In, K, Mg, Mn, Na and Rb was collected in counting vial and subjected to y-ray spectrometry. The column was then eluted with 5 ml of 0.3M HNO3, and the eluate containing As, Co, Cr, Cu, Fe, Hf, Mo, Sb, Zn and Zr was radioassayed. It took 7 minutes to acconplish the chemical separation prosedure. The activation products 239Np and 233pa from uranium and thorium, respectively, were determined using the y-ray spectrometric technique after chemical separation of these isotopes from the matrix. After dissolution of the irradiated target the oxidation state of neptunium was adjusted to Np (IV) by adding a suitable oxidizing agent such as KMnO 4. Then the solution was passed through an anion exchange column (Dowex 1X8; 400 mesh in R-C1 form) which retained 239Np and 233pa. The column was washed with 25 ml of 9M HC1 solution. Thereafter, 239Np and 233pa were eluted with 20 ml of 3M HF-9M HCI solution. Finally, 239Np and 233pa were coprecipitated with LaF 3 by the addition of 30 mg of La(III). To determine the chemical yield of RNAA procedures, the dissolved samples of non-irradiated nickel-alloy were spiked with the respective radiotracers of known activity. These were then sobjected to the separation procedures as described above. The yields were found to be in the range of 95-98%.
Gamma-ray spectrometry The y-ray spectra of the samples and standards were measured after appropriate cooling, for varying times ranging from 2 minutes to 24 hours using a 4k series 85 Canberra multichannel analyzer coupled with Ortec coaxial 30 cm 3 Ge(Li) derector. The system has a resolution of 2.0 keV for 1332.5 keV y-peak 6~ and peak/Compton ratio of 40:1. The data, transferred from MCA to the central computer facility, were processed using locally developed softwares. The y-spectrometry was repeated several times, to determine the half-life of each indicator radionuclide and to test its radiochemical purity.
J. H. ZMDI et al.: DETERMINATION OF TRACE IMPURITIESIN NICKEL-BASED ALLOY
Results and discussion
The optimized conditions and the nuclear data used for the determination of trace impurities in high purity nickel-based alloy were the same as described elsewhere 9 except the radiochemical separation procedure developed for this study. The short-lived indicator radionuclides, i.e., 28A1, 171Er, 27Mg, l16In, 165Dy, 56Mn, 72Ga, 24Na, 65Ni and 159Gd, were measured employing short irradiations and the relatively long-lived indicator radionuclides, i.e., 76As, 82Br, 47Ca, 140La, 153Sm' 239Np, 122Sb, 99Mo, 177Lu, 147Nd, 131Ba, 86Rb, 233pa, 51Cr, 169yb, 141Ce, 181Hf' 59Fe, 203Hg, 182gTa, 160Tb, 46Sc, 75Se, ll3gSn, 153Sm, 85Sr, 65Zn, 134Cs, 60Co and 152Eu, were measured employing long irradiations along with appropriate cooling times. The selection of the ),-peaks for the analysis of REE was the same as described elsewhere. 18 The photo-peaks 1596.5 keV, 145.5 keV, 531 keV, 103 keV, 1408 keV, 1178 keV, 94.6 keV, 184.5 keV, 283 keV, 308.3 keV and 208.4keV of 14~ 141Ce, 147Nd, 153Sm, 152Eu, 16~ 165Dy, 166mHo, 175yb, 171Er and 177mLu, respectively, were used for the determination of the respective elements. The ),-peaks of all the radionuclides, except 65Zn, 141Ce and 2~ were well resolved and free from any serious interference. The full energy peak areas of 1115.5 keV, 145.5 keV and 279.2 keV from 65Zn, 141Ce and 2~ were determined
after subtracting contributions from 1120.5 keV of 46Sc, 142.5 keV of 59Fe and 279.5 keV of 75Se, respectively. The reliability of our method was checked by analyzing British Chemical Standards; S. S. No. 310-1 (nimonic '90' alloy) and S. S. No. 456 (mild steel) employing the above mentioned chemical procedures. Our values are in fairly good agreement with the certified values as shown in Table 1. The concentration of 40 elements was determined and the analytical results, as averages of at least four determinations with standard deviations around mean values, are given in Table 2. The table shows that the composition of the alloy under investigation contains basically of Ni (56.8%), Fe (10.4%) and Cr (6.3%). Apart from the major components of the alloy, there are several other additions conferring certain properties to it. For example, manganese is used as a deoxidizer and makes the metal ductile. Its presence >1% increases the hardness but imparts brittleness, however, in the preseent case the manganese contents are only 0.3%.Vanadium makes the alloy very strong which does not break when subjected to severe shocks, on present case its concentration is 0.76%. Magnesium and sodium contents are 2.14% and 0.45%, respectively. The rest of the elements are present as trace impurities. The radiochemical separation procedures developed on this work have enhanced the sensitivity of the technique many folds as shown in Table 3.
Table 1. Analysis of British Chemical Standards (concentration in %*) Element
Cr Ti V Mn Fe Co Ni Cu AI Sn Sb
B.C.S./S.S.310/1 (Nimonic '90' Alloy) Our values Certified values 19.7 2.38 0.013 0.40 0.245 16.8 59.0 0.008 1.1 0.0010 0.0040
+ + + + + + + +_ + + +
0.2 0.30 0.002 0.03 0.018 0.2 0.2 0.002 0.1 0.0002 0.0007
19.4 + 0.1 2.43 + 0.29 0.35 + 0.01 0.250 + 0.013 17.0 + 0.1 58.6 + 0.1 1.06 + 0.03 -
B.C.S/S.S. 456(Mild Steel) Our values Certified values 0.018 0.0018 0.022 0.172 0.049 0.008 0.0050 0.008 0.0038 0.012
+ + + + + + + + + +
0.001 0.0004 0.002 0.012 0.004 0.001 0.0007 0.002 0.0005 0.003
(0.019) (0.002) 0.024 0.170 0.048 (0.007) (0.006) 0.008 (0.003) 0.011
_+ 0.001 + 0.004 + 0.003
+ 0.001 + 0.001
* Error quoted is the standard deviation. Values in parentheses are uncertified.
261
J. H. ZAIDIet al.: DETERMINATIONOF TRACEIMPURITIESINNICKEL-BASEDALLOY
T a b l e 2. Concentration of minor and trace elements in Ni-alloy (in Ixg/g)
Element
Concentration
AI As Ba Br Ca* Ce* Co Cr* Cs Dy Er Eu Fe* Ga Gd Hf Hg** In** La Lu
266.7 63.7 215 8.9 5.9 0.33 9.96 6.3 4.34 6.12 2.8 1.17 10.4 107.5 1.7 19.5 6 245 2.34 1.16
• _+ • +_ + + + + + + • • + + + • • + _+ •
21.6 7.0 11 0.8 0.2 0.2 1.06 0.8 0.41 0.61 0.45 0.10 0.49 9.4 0.4 1.6 1 28 0.22 0.10
Element
Concentration
Mg* Mn* Mo Na* Nd Ni* Rb** Sb Sc Se Sm Sn Sr Ta Tb Th U Yb V* Zn
2.14 0.296 9.8 0.45 16.2 56.8 24 2.2 0.13 0.16 0.23 0.25 371 1.21 1.24 18.3 4.3 2.48 0.76 93.4
_+ 0.19 + 0.0028 + i.0 + 0.03 + 1.5 +_ 7.6 + 3 + 0.2 + 0.01 + 0.02 • 0.02 • 0.02 + 30 • 0.11 _+ 0.10 • 1.9 + 0.3 + 0.22 • 0.8 • 34.3
* Concentration in %. ** Concentration in ng/g.
T a b l e 3. Comparison of derection limits achieved by RNAA with those of INAA (in ng/g)
Element AI* As Ba* Br Ca* Ce Co Cr* Cs Dy Er Eu Fe Ga Gd Hf Hg In La Lu
RNAA
INAA
Element
0.02 0.01 0.03 0.5 0.02 0.01 0.1 0.01 0.2 0.04 0.5 0.2 0.5 0.2 0.2 1 0.1 5 0.03 0.03
3 0.3 2 10 2 2 5 1 8 2 10 5 30 5 5 10 1 50 0.5 0.5
Mg Mn Mo Na Nd Ni Rb Sb Sc Se Sm Sn Sr* Ta Tb Th U Yb V* Zn
RNAA 1 0.5 0.1 0.1 0.2 5 0.2 1 0.1 1 0.5 0.2 I 0.02 0. I 0.01 0.01 0.1 0.01 2
INAA 10 2 10 1 2 30 5 10 2 10 10 4 20 0.5 5 2 2 2 5 20
INAA: tit r = 5 min to 10 h, t d = 30 min to 14 d, tm = 500 to 60000 s; RNAA: tit r = 2 min to 5 h, t a = 15 min to 7 d, t m = 300 to 30000 s; where %rr: irradiation time, ta: decay time, tin: measuring time. * Concentration in [ag/g.
Conclusions
RNAA procedure developed in the present work was successfully applied for the determination of trace impurities in nickel-based alloy.Since the Compton background was markedly reduced due to radiochemical
262
separations, therefore, the detection limits were considerably lowered for almost all the elements in general and short-lived indicator radionuclides in particular. The internal comparator and internal calibration methods inherit many uncertainties in the calculation of the absolute elemental concentrations,
J. H. ZAIDIet al.: DETERMINA'IIONOFTRACEIMPURITIESINNICKEL-BASEDALLOY
namely, relative efficiency calibration procedure (
