(Institute of Geophysical and Geochemical Exploration, Ministry of Geology and Mineral Resources)

Chen Yuanling Guan Bao

(Institute of Atomic Energy, China Academy of Sciences)

Chen Bingru Wang Yuqi Sun Jingxin

(Institute of High Energy Physics, Chinese Academy of Sciences)

Among modern analytical methods, neutron activation analysis is famous for its high sensitivity and high accuracy in the determination of many trace elements. In addition to high sensitivity and accuracy, neutron activation analysis has another advantage, which is non-destructive. In many cases, multiple elements can be determined simultaneously without chemical treatment. Pure instrumental neutron activation analysis at home and abroad can usually determine more than 25 elements. Neutron activation analysis technology has been applied to many aspects of geology, such as rock analysis, geological standard material analysis, discovery and identification of new minerals, analysis of precious samples collected on the moon or mountain climbing, rare earth composition analysis in geological theory research, geochemical sampling analysis and so on. , and completed a lot of high-quality analysis work, neutron activation analysis played an important role in the analysis of geoscience samples.

The fixed value analysis of the first batch of standard reference samples GSD 1-8 used for the first-level monitoring of regional geochemical scanning surface samples in China has aroused great concern in the activation analysis field. The Institute of Atomic Energy of China Academy of Sciences (hereinafter referred to as the Institute of Atomic Energy), the Institute of High Energy Physics of Chinese Academy of Sciences (hereinafter referred to as the Institute of High Energy Physics) and the Institute of Geophysical and Geochemical Exploration of the Ministry of Geology and Mineral Resources (hereinafter referred to as the Institute of Geophysical Exploration) determined 36 elements in GSD 1-8 by neutron activation analysis, and photographed short (several minutes) and long (10~20h ~ 20h. The Institute of Geophysical Exploration has developed an auxiliary software package (SPCSUP) to computerize the whole process of activation analysis data processing. The relative standard deviation of most elements is less than 65438 05%. All three laboratories use international standard reference materials GXR 1-6, AGV- 1, SY2-3, SRM- 1362A, etc. The reliability of the analysis method is verified, and the activation analysis results of most elements are in good agreement with the identification values.

In the preparation process of GSD 1-8, neutron activation analysis also undertakes particle size influence determination, vibration influence and uniformity inspection.

The study of neutron activation analysis method for GSD 1-8 samples shows the effectiveness of neutron activation analysis in the calibration of geochemical standard samples and its great potential in the analysis of geological samples. At the same time, many domestic experienced laboratories and several international laboratories have invested in the fixed value analysis of GSD 1-8 and adopted various analysis methods. In such a large-scale comparison work, the level of activation analysis in China has been well tested.

First, the method principle.

A variety of stable isotopes are irradiated by neutrons to generate artificial radioactive isotopes, which can measure the energy and intensity of gamma rays and analyze elements qualitatively and quantitatively. The basic formula of neutron activation analysis is as follows [1]:

Zhang Yujun on new methods of geological exploration.

Where: s (t)-radioactive intensity at time t;

Na-Afer Gadereau constant, 6.02×1023;

A refers to the gram atomic weight of the element to be measured;

P- sample weight;

W refers to the concentration of the element to be measured;

M- abundance of activated isotopes;

The number of target nuclei in the sample;

φ-neutron flux;

σ-activation cross section of target nucleus;

T-half-life of active nuclide;

T 1- irradiation time

Called saturation coefficient;

Td-cooling time;

E- detection coefficient.

Where w is an undetermined value, and other parameters are either known or measurable, so in principle, absolute neutron activation analysis can be carried out with the above formula; But in fact, it is quite difficult to measure the accurate values of φ, σ, E and other parameters, so the relative comparison method is usually adopted, and the activation formula is simplified as:

Zhang Yujun on new methods of geological exploration.

Where: s0 (w) —— the normalized value of the net peak area of the characteristic gamma-ray energy peak of the nuclide to be detected;

S0 (c) refers to the normalized value of the net peak area of the same energy peak of the same nuclide in the standard;

W refers to the content of the element to be measured in the sample;

C refers to the content of the same element in the standard.

Second, the analysis technology

In the analysis of GSD 1-8, the methods and technologies adopted by Institute of Atomic Energy and Institute of High Energy and Geophysical Exploration are basically the same, but there are differences in sample preparation, irradiation, measurement, pre-irradiation enrichment and data processing, which are briefly listed in Table 1:

1. Sample preparation and irradiation

Long exposure is the main method of neutron activation analysis, usually 10~20h ~ 20h, and the samples are packed with high-purity aluminum foil. Generally, after three days of unpacking and the first measurement, the activity of aluminum has decayed, and the operator will not be exposed to large doses of radiation. In short exposure, aluminum foil is not suitable for sample packaging because the purpose is to determine short-lived isotopes, but capacitor paper or cellophane is used instead. When irradiated by epithermal neutrons, the sample is wrapped in aluminum foil and placed in a small cylinder made of boron nitride. Its wall thickness is 1.5mm, and boron has a large absorption cross section for epithermal neutrons. Some elements (such as Ga, w, k, etc. ) absorbs epithermal neutrons, so epithermal neutron irradiation is beneficial to the determination of these elements. The irradiation was carried out on the heavy water reactor and two swimming pool reactors. The Institute of Geophysical Exploration and the Institute of Atomic Energy use their automatic and fast "rabbit-running" system to carry out short-fire tests, while the Institute of High Energy Physics carries out short-fire tests on small reactors, and then quickly transports them back to the laboratory for measurement.

Table 1 technical characteristics of three laboratory methods

All three laboratories use relative comparison method, so it is necessary to accompany the standard and sample in one irradiation tube at the same time. The adopted standards include single-element or synthetic multi-element chemical standards and several international standard reference materials (SRM). Chemical standards are lined with filter paper or plastic film; Powder standard reference materials and powder sample crystals were prepared by the same method before irradiation, dried and weighed.

2. Concentration before irradiation

In order to improve the representativeness of samples, suppress interference and improve the sensitivity of analysis of gold and rare earth elements, the pre-irradiation enrichment test of gold and rare earth elements was carried out. Before gold irradiation, 10g is used for enrichment and sampling, the ore is dissolved with 30% aqua regia, and then sucked by a Buchner funnel filled with activated carbon slurry. After carbonization, all the activated carbon paste cakes adsorbed with gold were wrapped in small bags made of high-purity aluminum foil for irradiation. Before irradiation, the rare earth elements were enriched and sampled as 1g, melted with sodium oxide at high temperature, precipitated with oxalic acid, ashed, dissolved with hydrochloric acid, and extracted with PMBP benzene extract. The obtained 10 ~ 15 ml rare earth solution was stored in a test tube. In the long exposure test of rare earth, take 1 ~ 4 ml of rare earth solution according to the total amount of rare earth in the sample, pour it into a porcelain crucible, add 10 layer of filter paper with the diameter of 10mm, evaporate in water bath, and wrap the filter paper with high-purity aluminum foil for exposure. In the short exposure test of rare earth, take 5 ~ 6 ml of rare earth solution, extract it with E 105 strong acid homogeneous cation exchange membrane for 48 hours, and wrap it with capacitor paper for later use. In order to calculate the relative transfer coefficient of rare earth elements, each treated sample is accompanied by its powder sample. Powder samples show accurate rare earth components, such as La, Sm, Ce, Eu, Yb, etc. Compared with the anomalies of these elements in the pretreated sample, the average migration coefficient of rare earth elements in the sample is obtained, and the content of other rare earth elements is calculated from this.

measure

The level of measuring equipment used in these three laboratories is similar. The measuring system of Geophysical Institute is Jupiter system, and the detection efficiency of Ge(Li) detector is 25%, and the resolution is 60Co 1.33 mev peak width at half peak.

Is 1.9 kev, and the resolution of HPGe detector is 122 kefwhm of 57Co, that is, 155eV. The measurement system of Atomic Energy Institute is SCORPIO—3000 system, the detection efficiency of Ge(Li) detector is 30%, and the resolution of 60Co 1.33MeV peak is 2.0keV. The black metering system is also a Scorpio-3000 system. The detection efficiency of germanium (lithium) detector is 28%, and the resolution is 65438±0.33 mev, the peak FWHM of 60Co.

The basic method of qualitative analysis of instrumental neutron activation analysis is "avoidance", which makes full use of time and energy to realize correct isotope identification. Not only different irradiation time, but also different cooling time are used to highlight nuclides with different half-lives. The analysis results of most elements are obtained through long-term exposure, and are measured 3-4 times after leaving the pile, and the cooling time is four days, two weeks, one month and three months respectively. In qualitative interpretation, try to choose the gamma-ray peaks with less mutual interference, with the peak with the least interference as the main peak, supplemented by one or more confirmation peaks. In qualitative interpretation, the size of activation cross section of each element and its possible concentration in ground samples should be fully considered. According to the nuclear parameters [2-4] and the actual geological conditions, the peak library and analysis library were re-edited. The main measurement parameters adopted by the three laboratories are listed in Table 2.

Modern high-resolution detector spectrometer can often correctly judge a variety of nuclides when using time parameters. For example, the energy of 1 10MAg's 657.744keV peak is close to that of 76As's 657.0keV peak, but their half-lives are quite different, the former is 253d, and the latter is only 26.3h After cooling for enough time, the peak of 1 10MAg's 657keV peak is enough. Another example is that the peak of1121.272kev of 182Ta interferes with the peak of1120.516kev of 46Sc, and their half-lives are very long, respectively.

Using high purity germanium detector can better distinguish γ peak in low energy region. For example, in the determination of Gd, the main peak is 97.432keV of 153Gd, which is interfered by 96.733keV of 182Ta, 153Sm and 75Se, of which 65438+. The interference of 153Gd can be avoided by cooling for a certain time, and the half-life of 75Se is 120d, so the interference of 183Ta and 75Se on 153Gd cannot be avoided by using the time parameters. Using high purity germanium detector can improve this problem, but when the content of 75Se is considerable, other supplementary means should be adopted to eliminate the interference of 75Se on 153Gd. For example, the interference value to be deducted is calculated from the ratio of the main peak of 75Se (136.0keV) to the peak of 96.733keV, or the rare earth elements are concentrated before irradiation activation.

Table 2 Main parameters of neutron activation analysis

sequential

sequential

There are many kinds of geological samples and complex matrices. The adaptability of various samples to neutron activation analysis is different, and the number of elements that can be analyzed is also different, mainly depending on how many weak peaks can be distinguished. The Institute of Geophysical Exploration understands the influence of adjacent peaks, Compton edges and Compton continuous lines on the resolution of weak peaks through experiments. A statistically sufficient spectral line is obtained with 137Cs, and its peak of 66 1.638keV is set as a strong peak, which is shifted by offset, attenuated in proportion, and then synthesized with the strong peak. Comparing the analysis results of independent peak and synthetic spectrum, the change of weak peak position is shown in figure 1. The experimental results show that:

(1) When the distance between two peaks is 10kev, that is, 5 times FWHM, the weak peak decays to 0. 1% of the strong peak amplitude, and it is still undisturbed.

(2) When the distance between two peaks is 5keV, that is, 2.5 times FWHM, the lowest distinguishable weak peak amplitude is 1% of the strong peak, and the error is 12%.

(3) When the two peaks are 2keV apart, that is, one FWHM, the distinguishable weak peak is 1/20 of the strong peak, and the error is 4.7%.

(4) When the distance between two peaks is 1keV, that is, 1/2FWHM, no matter what the amplitude ratio of the two peaks is, they are indistinguishable, and the analysis program regards them as one peak.

(5) When the weak peak is located on the Compton continuous line of the strong peak, the lowest resolvable weak peak is 0. 18% of the strong peak, and the error is 4.6%.

(6) When the weak peak is located at the Compton edge of the strong peak, the lowest resolvable weak peak is only 0.85% of the strong peak, and the error is 16.4%.

Fig. 1 interference test of strong peak to weak peak, schematic diagram of peak position and peak intensity change of weak peak.

This experiment shows that some strong radioactive isotopes, such as 24Na, 59Fe and 40Sc, often exist in geological samples after activation, and their gamma spectrum peaks and Compton edges will affect the analytical sensitivity and accuracy of low-content elements.

The dead time of the instrument is another main reason for the analysis error. No matter Scorpio-3000 system or Jupiter system, dead zone automatic compensation is adopted. No-load time is preset during measurement, and the experiment proves that the compensation of no-load time is carried out according to the following formula:

Zhang Yujun on new methods of geological exploration.

Where: tr-actual measurement time;

TL- preset measured survival time;

D- percentage of dead time.

Despite such dead time compensation, it is found that the compensation for counting loss is still insufficient. Moreover, the degree of counting loss is also related to the size of the counting rate itself. The weaker the counting rate, the more serious the loss or insufficient compensation is. The 1460keV peak at 40K is a stable peak in the environmental background. Figure 2 shows the change of the net peak area of the peak under different dead zone conditions and when the preset measurement activity time is 500 seconds.

In order to verify that the compensation of dead time is unequal for peaks with different intensities, we take 60Co source and 137Cs source, and adjust the source distance so that the peak of 61kev, the peak of 60Co 1332keV and the peak of 40K1kev are 6137cs. In other words, the smaller the peak intensity, the more serious the under-compensation. Therefore, in order to ensure the measurement accuracy of weak peaks and control the source distance, the measurement dead time of the strongest samples should not exceed 10%.

Fig. 2 Variation of 40K peak area with dead time under environmental background.

Neutron activation analysis accuracy of elements with large activation cross section is quite high. However, the analysis accuracy of those elements with small activation cross section or too low content will decrease. As can be seen from the above analysis, the main error sources of weak peak measurement are 4:

σ1-statistical error, which is assumed to be10% in weak peak measurement;

σ 2 —— the error caused by dead zone under-compensation, taking 5%;

σ 3 —— The error caused by strong peak interference is set at 5%;

σ 4 —— Error from other sources, such as weighing samples, neutron flow gradient and changes in measurement geometry, is also set at 5%.

Then the total measurement error σ is expressed as follows:

Zhang Yujun on new methods of geological exploration.

Substitute σ 1, σ2, σ3, σ4,

Zhang Yujun on new methods of geological exploration.

This estimate is consistent with the actual situation, and the analysis accuracy of weak peaks is usually within 15%.

4. Data processing

The spectral analysis program of SCORPIO-3000 system is Scorpio/Spectran, and the spectral subroutine of Jupiter system is SPECT-RAN-F. The principle is the same, and the analysis result is the activity ratio of each isotope in each sample, that is, the activity in unit sample volume (volume or weight):

Zhang Yujun on new methods of geological exploration.

Among them:

Zhang Yujun on new methods of geological exploration.

λ is the attenuation constant, λ = log (2)/t;

Tc-is the measured time, not the survival time;

C refers to the correction when the isotope with short half-life decays obviously in the measurement process;

A- peak area;

Td-cooling time;

T isotope half-life;

T 1- field acquisition time;

Ω —— the yield of gamma rays;

V- sample size;

ε-detection efficiency.

For a given peak of a given isotope, the detection efficiency ε of the same instrument is constant. For comparative activation analysis, ε value has no practical significance, because it is always compared with the same energy peak of standard and sample. But this parameter is needed when the program is executed, so the virtual calibration method is adopted to input all the efficiency values of different energies into 1, which simplifies the process of efficiency calibration.

In order to calculate the element content, a series of complex and error-prone data processing work is needed, and geological samples are often activated in batches and repeated many times. In order to reduce errors and simplify data processing, the Institute of Geophysical Exploration developed a supplementary software package SPCSUP (SPECTRAN-F Supplementary Software Package) on the basis of mastering Spectran-F basically, and realized the computerization of the whole data processing process.

SPCSUP consists of 10 program function blocks:

* ACFL: calculate the activation strength, which is used to select the best irradiation and measurement time, formulate the element formula of artificial standard, and estimate the safety during measurement or unpacking.

* TRFL: Simplified spectrum transmission operation and reduced the number of keyboard questions and answers to four.

* CUFL: Call and combine calculation program blocks, preset initial parameters, and provide files needed for calculation.

* STEDIT: Prepare standard library files.

* BMFL: calculation content working curve. When the value standard number is ≥3, the least square method is used, and when the value standard number is ≤2, the slope or slope average method is used.

The general form of univariate linear regression equation is:

Zhang Yujun on new methods of geological exploration.

According to the content and radioactivity of more than three valuable standards, the slope m and intercept b of the tropic of cancer can be obtained. When calculating, add a set of numbers w0 = 0 and i0 = 0.

Zhang Yujun on new methods of geological exploration.

*LTFL: List the comparison results and calculate the element transfer coefficient. The list file summarizes the quantitative analysis results of1-12 spectral lines into a file that only takes up 5 blocks, and the space occupied by12 spectral lines is at least 35×12 = 420 blocks.

* CNFL: content calculation, which can calculate the content of all elements in1-12 samples at one time. BMFL and CNFL designed the automatic deduction of uranium fission effect.

* AVFL: calculate the results of several measurements or exposures as the average value and standard deviation s.

Zhang Yujun on new methods of geological exploration.

* FRFL: list of general reports.

* LIRP: check the multiple exposure analysis values of each element, queue up and eliminate bad values, and print out the final report. The elimination of outliers is based on Dixon test standard, which is also programmed.

The relationship between the above functional blocks and the logical relationship with the spectral analysis software SPECTRAN-F are shown in Figure 3.

Fig. 3 Block diagram of neutron activation analysis software package

The main features of SPCSUP auxiliary software package design are: ① adopting virtual efficiency scale; ② Using SPCLST to realize the call of quantitative analysis results; (3) editing the standard data into a callable file; ④ All the intermediate results are saved in file form; ⑤ Batch dynamic combination files; ⑥ realize the degree entrance by batch calling editing; ⑦ Use the design skills of computing operation software to simplify the call of function blocks and optimize the man-machine relationship.

The application effect of this software package is obvious. Compared with semi-manual calculation, the data processing speed is increased by more than ten times, especially the error of manual calculation is avoided, the data is easy to look up, and the results are easy to save and copy.

Three. Results and discussion

As the first batch of standard reference samples GSD 1-8 series of stream sediments used for the first-level monitoring of regional geochemical scanning surface samples in China, after three years of development, they passed the ministerial appraisal in July of 1983, and put forward recommended values or reference values (hereinafter referred to as fixed values) for 54 elements [6]. Thousands of analysts from China 4 1 Laboratory contributed to this work. The Canadian Geological Survey and the French Institute of Geology and Mineral Resources also provided analytical data. There are as many as 26 analytical methods 1 1, and the workload of each method is listed in Table 3.

Table 3 workload comparison of various methods

Three neutron activation laboratories submitted the analysis results of 36 elements, as shown in Table 4. As can be seen from Table 3, the original data submitted by neutron activation and mass spectrometry account for 7% of the total data. Although it ranks seventh in 1 1 method, the number of elements submitted by neutron activation analysis (36) accounts for 53% of the total number of 68 elements. Among the 36 elements, there are 3 1 (silver, arsenic, barium, cerium, cobalt, chromium, cesium, dysprosium, europium, iron, gadolinium, hafnium, potassium, lanthanum, lutetium, manganese, sodium, neodymium, rubidium, antimony, scandium and rubidium. The other five elements were analyzed only in an activation laboratory: Au and Ho were determined by irradiation enrichment method in Geophysical Institute, Ga was determined by epithermal neutron (boron nitride coating) irradiation method in atomic energy, and Br and Ni were determined by high energy research institute in swimming pool reactor.

Among the 36 elements, Au and Br have not been given fixed values at the appraisal meeting held in 1983, and 8 samples are all low, which is out of the original data set involved in the calculation of fixed values. The neutron activation data of the other 33 elements involved in the fixed value calculation are all accurate, and most of them are located in the middle of the original data set. These 33 elements account for 6 1% of all 54 elements. It can be seen that neutron activation has obvious advantages in calculating the number of elements in fixed value among various methods. Especially, As, Ce, Cs, Dy, Eu, Gd, Hf, Ho, La, Lu, Nd, Rb, Sc, Sm, Ta, Tb, Tm, u, Yb and other elements in neutron activation analysis are more difficult, so neutron activation is more important. In quantity and quality, the fixed value analysis level of GSD 1-8 series in three neutron activation laboratories can be compared with similar work abroad (33 elements were determined by SY2-3 in Canada, 28 elements by JB- 1 and JG- 1 in Japan, and 35 elements by GXR series in the United States). Moreover, neutron activation also plays a very good role in particle size and uniformity test.

(1) neutron activation analysis results have good precision and accuracy. Taking the activation analysis results of 33 elements submitted by geophysical exploration as an example, the accuracy is as follows: after repeated determination (10 ~ 20), the relative standard deviation of 28 elements is less than15%; The accuracy is as follows: 80.7% of all activation results deviate from the set value within 20%; If only Ge(Li) results are taken, excluding HPGe and pre-irradiation enrichment analysis results, 1 154 in 2 1 comparable data is less than 10%, accounting for 73% of 2 1 1 total; The deviation is less than 20% 183, accounting for 86.7%. One of the reasons for the difference is that the extraction rate of heavy rare earth before rare earth enrichment is lower than that of light rare earth. Fig. 4 shows the comparison between the data of Ge(Li) detector 2 1 1 and the fixed value, and the results of frequency statistics on the relative deviation. This diagram can directly show the accuracy of neutron activation analysis.

Fig. 4 Statistical table of relative deviation frequency between germanium (lithium) analysis results and fixed values in Geophysical Institute.

(2) The analysis results submitted by the three neutron activation laboratories are generally consistent, but by carefully comparing the elements, it will be found that there is a certain systematic deviation trend for some elements in each laboratory. For example, the determination of Yb is as low as 10% ~ 19% in Institute of High Energy and 7% ~ 20% in Institute of Geophysical Exploration. Only after careful experimental study can we make further accurate evaluation, but as far as equipment conditions and measurement technology are concerned, it can be safely considered that the standards adopted in activation analysis are the key factors affecting accuracy.

(3) 3) All activation data of Sb are higher than the fixed value, and the three activation laboratories are close to each other. The activation analysis mainly determined the 1690keV peak of 124Sb, and this energy peak had no interference. The Institute of Geophysical Exploration also measured the 564keV peak of 122Sb, and the results were consistent. GSD 1-8 was also determined by neutron activation method in May, 1984. The analytical results of antimony were 0.28ppm, 0.75ppm, 6.6ppm, 2. 1.5 ppm, 4. 1ppm, 1.84ppm and 2.96ppm, respectively. It can be preliminarily considered that the result of neutron activation of Sb is accurate.

(4) Table 4 also lists the results of 1 1 elements, such as Ce, Gd, Hf, Ho, Lu, Nd, Sm, Ta, Tm, W, Yb, etc. , used for geophysical exploration. This probe is helpful to make full use of the energy peak in the low energy region, in which Ho is only the result of HPGe detector, and the rest is 10 elements.

(5) Neutron activation analysis before irradiation can reduce the detection limit, improve the anti-interference ability and improve the sample representativeness. For example, after enrichment before irradiation, the detection limit of Au decreased from 10-8 to10-1. In addition, the enrichment of rare earth before irradiation also detected Ho which was not detected by instrumental neutron activation analysis.

(6) Taking GSD- 1, 2 and 6 as examples, the curve of the rare earth model diagram (Figure 5) made by the average results of three neutron activation laboratories (Table 5) is smooth, which shows the reliability of neutron activation in rare earth element analysis. The rare earth model curve of GSD-2 has three obvious characteristics: high rare earth content, serious Eu loss and high heavy rare earth content; This is consistent with the explanation that the sampling area of this sample belongs to the young granite development area. It is further proved that neutron activation analysis is a favorable means for rare earth composition analysis in geological theory research.

Table 5 Average content of rare earth elements and model curve values

Fig. 5 GSD- 1, 2,6 rare earth model diagram

(7) Uniformity is a necessary condition for standard samples. Neutron activation completed the uniformity test of GSD-2, and some uniformity test results are listed in Table 6.

Table 6 Partial results of uniformity test of GSD-2 neutron activation analysis

Table 6 shows that the measured values of F0.05 and t0.05 are less than their critical values. This shows that when the significance is 5%, there is no statistically significant difference between the total variation and the analysis variation, and between the group A average representing the population and the group B average representing the sub-samples. Although the sampling amount of neutron activation analysis is only tens of milligrams, no obvious sampling deterioration is found. It is further proved that the uniformity of the sample is good.

Table 7 neutron activation analysis of GSD-2 in different particle sizes of different properties of elements.

Fig. 6 Jupiter multichannel spectrometer system

(8) The content changes of elements in different particle sizes of GSD-2 were determined by neutron activation method. Table 7 lists the distribution of eight elements, such as Hf, Ta, Ce, Rb, Cs, Co, Zn and Fe in different particle parts of GSD-2 sample. The table shows that Rb, Cs, Co, Zn, Fe and other elements mainly exist in easily broken and weathered minerals and rock-forming minerals [5], which are enriched in the fine fraction and change gently in the coarse fraction, so it can be judged that these elements have good uniformity in the sample. Hf, Ta, Ce (representing rare earth) and other elements mainly occur in zircon, monazite and other stable and wear-resistant accessory minerals or trace minerals, but are enriched in coarse particle size, so the sampling error of such elements can not be ignored.

Neutron activation also analyzes the experimental samples that simulate the influence of transportation vibration on sample uniformity, and the results are listed in Table 8. Generally speaking, it is not found that the uniformity of the samples is obviously worse after being transported by car.

Table 8 Experimental results of GSD-2 neutron activation analysis to check the influence of vibration on uniformity

Thank you. The enrichment method of rare earth elements used in this paper was formulated by Comrade Yu of Shaanxi Bureau of Geology and Mineral Resources. Zhang Yuanji, Zhao Meizhuo and Shi Jianwen of Geophysical Research Institute all participated in the analysis. I want to thank all of you.

refer to

De soete D., Gij Bels R. and Hoste J.: Neutron activation analysis, 1972.

[2] Crouthamel C.E.: Applied Gamma Ray Spectroscopy, 2nd Edition, 1970.

[3]Heath r. l .:γ -ray spectrum catalogue Ge(Li) and Si(Li) spectrometry, Vol. 1 and Vol. 2 1975.

[4] radionuclide data sheet. Beijing: Atomic Energy Press, 1925.

Yan et al. Preparation of geochemical standard samples of stream sediments, geophysical and geochemical exploration, 198 1, 5(6).

[6] Xie et al.: Geographic Standard Newsletter, Volume IX,No. 1, 1985.

Neutron activation analysis technology of reactor

China geochemical standard sample GSD1-8

Zhang Li Yujun Xingbin Songlin Mountain

(Institute of Geophysical and Geochemical Exploration, Ministry of Geology and Mineral Resources, China)

Chen Yuanling Guan Bao

(Institute of Atomic Energy, Academia Sinica)

Sun Jinxin, Chyi Yu, Wang Ru, Chen Bing

(Institute of High Energy Physics, Academia Sinica)

The contents of 36 elements in geochemical standard samples of China Ministry of Geology and Mineral Resources were determined by neutron activation analysis. Three main variants of this technique, namely instrument, epithermal neutron and preirradiation neutron activation analysis (INAA, ENAA and PNAA), were used in the systematic study of GSD1-8 samples by three laboratories. Long irradiation and short irradiation as well as Ge(Li) and HPGe detectors are used. A supplementary data processing software package has been developed.

Note: Table 4 is omitted. See Table 2 of the paper "The Important Role of Reactor Neutron Activation Analysis in the Development of the First Batch of GSD 1-8 Geochemical Standard Samples". When editing this anthology, in order to make this work have a complete historical record, Figure 6 (Jupiter multichannel spectrometer system) is supplemented.

Collection of exploration geophysics and exploration geochemistry, episode 4.