26.04.2016 г.

  На главную раздела "Научные работы"



O. Evdokimova1, M. Zaitceva2, N. Pechishcheva1, K. Shunyaev1, A. Pupyshev2,
S. Noritcin2

1Institute of metallurgy of the Ural branch of the Russian Academy of Sciences,
620016, Yekaterinburg, Russian Federation
2The Ural federal university, 620002, Yekaterinburg, Russian Federation.
Correspondence: evdokimova.olgav@gmail.com


          The sample preparation of tungsten concentrates by fusion for X-ray fluorescence analysis of tungsten, calcium, silicon, manganese and sulfur was developed. The conditions of sample preparation by fusion with investigated fluxes were optimized. A thermodynamic modeling of fusion process shows there are no losses of tungsten, calcium, silicon, manganese at the experimental conditions and the losses of sulfur are minimal or absent.


          Industrial tungsten ores contain only about 0.5 - 1% of W thus there is an enriching of ore to produce tungsten concentrates. To control the enrichment process it is necessary to have the rapid methods of chemical analysis allowing determining the content of the main components of ores and concentrates with high accuracy. Tungsten in ores is contained in the form of mixture of different wolframites, for example FeWO4, MnWO4, CaWO4 (1). The typical components contents in concentrates is the following: WO3 (20-70 % wt.); MnO (0,1-20 % wt.), CaO (0,5-10 % wt.), SiO2 (1-10 % wt.), Fe3O4 (20-30 % wt.), P (0,05-2 % wt.); S (0,05-4 % wt.) and other. A number of analytical methods are available for the analysis of tungsten concentrates: spectrophotometry, titrimetry, the weighting method (2). Most of them are the labor-intensive and time-consuming. The XRF is the most popular technique especially when a large number of samples of similar type are to be analyzed. Moreover, this does not involve solution of the sample.

          The optimization of sample preparation is the most important stage of any methods of chemical analysis. This stage determines the quality of the analysis, the reproducibility and accuracy of the result. The number of publications about to the development of XRF techniques for tungsten ores and concentrates is limited. Most of these techniques allow determining the content of only a few components, such as tungsten only (3-5) or tungsten and manganese (6). Such investigations that allow extending the number of the determined components at the same time are demanded for metallurgical enterprises.

          The literature describes various methods of sample preparation of tungsten ore materials for XRF analysis. Among them pressing technique (3, 7) and fusion technique with fluxes (4-6, 8-10) are the most common. Fusion technique has a number of advantages compared to pressing technique, for example, it eliminates the dispersion degree effect of the materials and the effects of mineralogical composition, it allows introducing additional substances such as absorbents or internal standards to reduce or compensate of matrix effects. Also the possibility of sample diluting reduces the influence of the matrix. The fusion techniques for tungsten ores and concentrates with fluxes described in the works (4-6, 8-10) require the use of muffle furnaces and expensive Pt-Au crucibles. Necessity of addition LiF or NaF to reduce the melt viscosity and increase the wettability of the crucibles by melt makes it impossible for determination of silicon in tungsten concentrates because of formation of volatile silicon compounds.

          The challenge of the work was to development the sample preparation technique of tungsten concentrates by fusion with fluxes to the XRF determination of tungsten, calcium, silicon, manganese and sulfur. In this work fusion of samples with fluxes was carried out in a furnace with resistive heating using graphite discs. It allowed us not to use expensive equipment to fusion.


          Certificated reference material of tungsten concentrates (CRM -1712 and CRM-1710) and samples of tungsten concentrates, the composition of which was determined by ICP AES, were used for the procedure development (Table 1). X-ray fluorescence analysis was made by X-ray spectrometer «S4 Explorer» («Bruker», Germany). Fusion of tungsten concentrates was carried out on the graphite discs of diameter 40 mm in the furnace "MAX-2M".

Table 1. Composition of samples used for procedure development
Пример изображения


          Sodium tetraborate, lithium tetraborate and metaborate are most often used as a flux for fusion. Lithium compounds as a flux is preferable as it absorbs the longwave radiation and can reduce the detection limit. Metaborates have a lower melting point than tetraborates, and possess a strong alkaline properties, thus metaborates is better suited for fusion of samples with a large amount of acidic oxides, e.g. SiO2. Tetraborates are preferable for fusion of alkali oxides such as CaO (11).

          It was considered Li2B4O7, Li2B4O7 + LiBO2 (1:1), Li2B4O7 + Na2B4O7 (1:1) as a flux for fusion of tungsten concentrates; LiNO3, NaNO3, Li2CO3, Na2CO3 as an additive to reduce the melt viscosity. All fluxes were calcined at 400° C.

          Two tablet emitters of each sample on graphite disc were measured 3 times, the intensity of the elements was fixed. As a result of experiments on the choice of the optimum additive it was found values of the spectral lines intensity for all investigation elements for using nitrate and sodium carbonate as additives was lower than for nitrate and lithium carbonate. Moreover the values of the relative standard deviation (RSD) of results for these additives nitrate and sodium carbonate are high. Results of experiments on the choice of the optimum additive are summarized in Table 2.

Table 2. Values of the spectral lines intensity of the elements and RSD
of analysis results for different additives (n = 2) (mass of tungsten concentrate - 1 g,
Li2B4O7 -2.5 g, additives - 0.25 g)
Пример изображения

          The lowest RSD values are observed for lithium carbonate, moreover this additive promotes uniform spreading of melt on the disc, so it was chosen for further work. At a fixed mass flux of 2.5 g and sample of 1 g, for fluxes Li2B4O7, (Li2B4O7 + LiBO2), (Li2B4O7 + Na2B4O7) and additives Li2CO3 the optimal weight of additives, time and temperature of fusion were determined while simultaneously varying these parameters. Time of fusion was varied in the range of 10-15 min, the temperature of fusion - from 950 to 1000 ° C, additive mass - from 0.25 to 0.6 g (for Li2B4O7 + Na2B4O7) or 0.4-0.6 g (for Li2B4O7, Li2B4O7 + LiBO2).

          The intensity of the spectral lines was used as the parameter of optimization for Si and S, because for low concentrations it is necessary to fix a maximum value of intensity. As for macro components W, Ca, Mn, whose spectral lines intensity is high enough, RSD values were used as the parameter of optimization. The optimal conditions of fusion for each flux are shown in Table 3.

Table 3. The optimal conditions of fusion for fluxes
Пример изображения

          Table: RSD,% - the total value of the RSD,% for all elements; I - average intensities of all elements, kimp/sec.

          For further sample preparation by fusion flux Li2B4O7 was chosen, it promotes minimum error of measurement of the spectral lines intensity for macro-component because uniformity of the surface of the melt is better.

          As a result, the following optimal conditions for sample preparation by fusion of tungsten concentrates on a graphite discs (for the ratio of the mass concentrate: mass flux = 1 g: 2.5 g) are:

          • Mass of additives Li2CO3 - 0,375 g;
          • The temperature of fusion - 975 ° C;
          • Time of fusion - 12.5 min.


          Losses of volatile components such as sulfur, phosphorus are possible when fusion of materials is occurred at high temperatures. In order to prove that the losses of volatile components are absent or minimal for selected experimental conditions the thermodynamic modeling of the fusion was performed by HSC software.

          Conditions and modeling assumptions:

          - The temperature range from 25 to 1025 ° C;
          - the ratio sample:air (inert atmosphere of nitrogen; sample: air =1:1, 1:10 and 1:100);
          - Composition of the sample: 1 g of tungsten concentrate, 2.5g of lithium tetraborate, 0.5 g of lithium carbonate. The average composition of tungsten concentrates was used for the calculations (Table 1).

          The conclusions according to the thermodynamic modeling results was following:

          a) Loss of tungsten, calcium, manganese, silicon in all the cases is not observed.

          b) The gaseous of sulfur compounds as SO2 are observed in a nitrogen atmosphere (in the absence of air) from T = 625 K, as well as a bit amounts as SO2 and SO3 in the presence of air excess (probe: air = 1:100) from 825 K.

          There is no losses of macro component and sulfur gaseous are minimal or absent for all the considered atmospheres under experimental conditions (950 ° -1000 ° C). For example, Fig. 1 shows the content of the compounds determinated components of the temperature at a ratio sample: air =1:10.


          Previously sample preparation technique of tungsten concentrates was optimized by pressing of powdered samples with a binder wax into pellet: 4 g of tungsten concentrate, 2 minutes of grinding time, the binder - wax (sample:wax=5:1), compaction pressure - 300 kgf/cm2(7). Comparison of XRF analysis results for tungsten concentrate samples using preparation by fusion with flux and pressing of powdered samples with a binder are shown in Table 4. In general, RSD values for fusing technique are less than for pressing.

Пример изображения
Fig. 1. Behavior of elements during fusion of tungsten concentrate in air
for a ratio of probe: air (1:10): a - W; b - Ca; c - Mn; e- Si; d - S (Modeling results).

Table 4. Analysis results of tungsten concentrate sample using different techniques of preparation, n = 12
Пример изображения

          It's necessary 4 g of sample for making pellet by pressing technique (7), while for fusion only 1 g of sample is enough. Moreover, using fusion of concentrate with a flux sample evenly spreads on graphite discs and hence the distribution of components in the fused pellet is equally. As for pressing technique (7), the influences of grains and different size particles are not excluded. Moreover the components in pressed pellet may be distributed unevenly, it can cause the significant errors of element determining.


          The conditions of sample preparation of tungsten concentrates by fusion with fluxes on the graphite discs in the furnace "MAX-2M" for XRF analysis were optimized. The thermodynamic modeling of fusion process shown there are no losses of macro component at the experimental conditions and the losses of micro component are minimal or absent. The RSD values of element determination results for fusing technique are less than for pressing.

          The study is supported by Program of UD RAS, project No 12-P-3-1004 and carried out using equipment of Collective Center "Ural-M".


          1. Zelikman A. N.: Metallurgiya tugoplavkikh redkikh metallov (Metallurgy of Refractory Rare-earth Metals). Moscow, Metallurgy, 1986 (in Russian).

          2. National standard 11884.10. Tungsten concentrate. Moscow, Publishing House Izdadelstvo standartov, 1978 (in Russian).

          3. Chacharkar M.P., Sathe M.D., Lal M.: Determination of tungsten in ores by energy dispersive X-ray fluorescence analysis. J. Radioanal.Nucl.Chem., Letters, 1987, 119 (1), 29-35.

          4. Harris A.M., Hingle D.N.: Determination of tungsten in ores, concentrates and plant process materials using internal standard X-ray fluorescence spectrophotometry. X-Ray Spectrometry, 1979, 8(3), 129-134.

          5. Zhao H., Zhang W.: Determination of tungsten and molybdenum in minerals by fusion beads X-ray fluorescence spectrometry. Metallurgical Analysis, 2007, 27(5), 25-29.

          6. Chen P.: Determination of tungsten, manganese, iron and niobium in pure tungsten mineral and tungsten ores using X-Ray fluorescence spectrometry. Chinese Journal of Analytical Chemistry, 1985, (7).

          7. Evdokimova O., Zaytseva P., Pechischeva N., Shunyaev K., Pupyshev A.: Optimization of sample preparation for XRF Determination of W, Ca, Si, Mn, P and S in tungsten concentrates. 11th Israeli-Russian bi-national workshop «The optimization of composition, structure and properties of metals, oxides, composites, nano- and amorphous materials», Chernogolovka, 2012.

          8. Pu Xu-li, Wang Hong-hui, Ye Shu-ai, Qing-mu Dong, Zou Jian-long, Yang Mink-kun: Determination of major and minor components in tungsten concentrates by wavelength dispersive X-ray fluorescence spectrometry. Rock and mineral analysis, 2010, 29 (2), 143-147.

          9. Wang Z., Niu S., Deng H.: A simple method for the determination of tungsten ores by XRF spectrometer. Spectroscopy and spectral analysis, 1999, 19 (1), 93-94.

          10. Srivastava S.C., Bhaisare S.R., Wagh D.N.: Analysis of tungsten in low grade ores and geological samples. Bull. Mater. Sci., 1996, 19(2), 331-343.

          11. Ehrhardt H., ed.: X-Ray Fluorescence Analysis: Application in Industrial Laboratories. Moscow, Metallurgiya, 1985 [Russian translation from German].

Добавить комментарий Сообщение модератору

Защитный код