Introduction

Materials containing Ge and/or In are of all types: from minerals and concentrates, by-products of large-scale metallurgy, to municipal waste and end-of-life products. These materials differ primarily in their chemical composition and origin. The contents of Ge and In in them can reach up to 10 and 6 wt%, respectively, although they are often below 1 wt%. These products are the result of various hydro- and pyrometallurgical processes for obtaining other metals such as zinc, lead and copper1,2,3,4. Table 1 shows intermediates with known chemical compositions, containing Ge and In. The germanium-indium drosses (Ge-In-D) discussed in this work are a by-products of zinc and lead metallurgy, hence the technological processes related to these metals will be presented later.

Table 1 Chemical compositions of germanium and indium by-products from heavy metal metallurgy.
Full size table

Pyro- or hydrometallurgical methods are used to produce zinc from primary or secondary sources. Technologies for zinc metallurgical production are divided into: - horizontal batch process, - in a vertical retort, - Imperial Smelting Process (ISP, Ltd), - in an electric furnace10. A typical ISP consists of several stages: (1) Preparation of the charge (briquette/sinter), (2) Smelting of Zn and Pb in a shaft furnace (smelting furnace, ISF), (3) Rectification of Zn and Cd.

Zn-Pb concentrates are mixed with residues from previous process and fine sinter fractions, with secondary material and other technological materials. The mixture is fed to a sintering machine with upper/lower draft and ignited. During sintering, the material is transported through a series of air boxes. As a result of sintering, sulfides are oxidized and mainly SO2 is released, and the heat generated is sufficient to melt and sinter the material, resulting in a sinter – Zn and Pb concentrate devoid of most of the sulfur, at the same time porous enough to be permeable to gases, and mechanically durable. SO2 is sucked in by the suction nozzle and sent for disposal into H2SO411,12.

Sintering takes place in an shaft furnace. Hot sinter (~ 45% ZnO, ~ 20% PbO) together with hot metallurgical coke and fluxes are loaded into the furnace through a bell system. The oxidizer, i.e. air or air enriched with oxygen, is blown through nozzles into the furnace. The coke oxidation reaction occurs, which, under appropriate conditions, favors the Boudouard reaction, which is mostly responsible for the reduction reactions of metal oxides in the sinter. High temperature (up to 1300 °C) and reducing conditions cause the melting of metals, of which zinc, due to its high vapor pressure, distills along with the furnace exhaust gases. The furnace gases pass through a spray condenser, in which a downpour of liquid lead (spraying) cools the gases to approximately 450–500 °C. The zinc condenses in the lead. This treatment prevents Zn reoxidation (cooling of Zn vapor occurs without the access of the oxidant). The resulting Zn-Pb alloy separates as a result of its free cooling: Zn, being lighter, forms the upper layer in the liquefaction bath, and the heavier lead forms the lower one. Some of Pb is returned to the condenser11,12.

Raw zinc, due to a certain amount of impurities, especially Cd, Cu and Pb, should be refined in rectification columns. The columns consist of refractory shelves through which metal vapors flow. There is an overflow hole at the bottom of each box. Adjacent boxes are positioned relative to each other so that the liquid metal or its vapors must flow from one box to the other in a zigzag manner. The lower parts of the columns are heated, and the upper parts are not - rectification takes place in the dephlegmator. Low-boiling components are rectified, while those with higher temperatures fall to the bottom of the column. In the zinc column, the rectification of Zn takes place, which evaporates together with Cd, and in the cadmium column, the separation between Zn and Cd takes place. The distillate from the first column is directly directed to the second column. The metal from the bottom of the second column is high-grade zinc (SHG)12,13,14,15. During the production of primary zinc in the New Jersey process, various types of by-products are produced, which may include Ge and In (Table 1)1,4,16.

The goal of this article was to characterize, in terms of material, germanium-indium drosses originating from heavy metals metallurgy. Although the goal may sound prosaic, the literature review shows that there is still no reliable and complete material characterization of Ge-In-D. The literature data contradict each other with the results of the chemical composition of the Ge-In-D, which may be due to several reasons: - the input from which the Ge-In-D are made is variable (oxide or sulfide charge, which determine the composition of the concentrate), - technological parameters vary in different plants, - various scientific methods may give different results. Almost always, the characterization of Ge-In-D are limited only to the analysis of the quantitative chemical composition, which ignores the important issue of microscopic analysis. Due to the above, a number of studies were carried out to characterize Ge-In-D.

Methods

The research material are germanium-indium drosses derived from the zinc distillation process as a solid residue. This material was tested using several research methods: - granulometric analysis; - microscopic analyzes (optical and electron), - analyzes of quantitative chemical composition.

Due to the nature of the research material, which is characterized by different grain sizes, and the insinuation that some fractions may be richer in Ge and/or In, a sieve analysis of the research material was performed. Sieve analysis was performed based on sieving according to the polish standard PN-90/H-04933, with the difference that the sieving was performed manually (not automated)17. Sieve analysis was performed for raw and ground research material. The ground material was prepared by grinding the raw material in a ring-roll mill (TEST-LAB-09, „Eko-Lab”) for 4 min.

Samples for sieving (dry sample, drying time 24 h, 110 °C, electric dryer SLN 115 STD, Pol-Eko-Aparatura Sp.J.) were sieved on a set of sieves (VEB Metallverbei Neustadt/Orla, Kombinat NAGEMA) in the sieve size range from 2.5 to 0.040 mm. Weighing samples for screening in the amount of 100 g ± 0.1 g and the fractions after screening were weighed with an accuracy of 0.1 g (WPS-360/C, Rad-Wag). Sieving time was 20 min. After sieving, each of the separated fractions was weighed. The sieving loss was added during recording to the mass of the undersize fraction collected in the bottom17. Sieving for each material was performed nine times using the same tools and under the same laboratory conditions. In subsequent studies, the separated fractions constituted research material for determining their chemical composition, as well as the shape of grains, etc.

Microscopic investigation was performed. For this purpose, metallurgical specimens of materials and their fractions were performed (EpoFix resin, EpoFix hardener, Struers). Then, the sections were ground (# 4000, MR54, Struers) and a conductive layer of carbon was applied to them using various sputtering machines (208 carbon, Cressington Carbon Coater or Q150TE, QUORUM).

The specimens were used for microscopic and morphological analyses: i.e. optical microscopy (EPIPHOT 200 metallurgical microscope, Nikon), scanning electron microscopy with an energy dispersive spectroscopy system (SEM-EDS) and element mapping (field emission gap, SU-70, Hitachi), and also an electron microprobe equipped with optical microscopes for reflected and transmitted light (JEOL Super Probe 8230). These analyzes allowed not only the assessment and determination of the morphology, but also the elemental composition of the materials.

Fractions of the material were additionally examined in terms of chemical composition with an electron microprobe JEOL SuperProbe JXA-8230 (EPMA) at the Laboratory of Critical Elements, AGH University of Science and Technology, Kraków equipped with X-ray wavelength spectrometers (WDS) and an X-ray energy dispersive spectrometer (EDS). The EPMA was operated in the wavelength-dispersion mode, at the accelerating voltage of 20 kV and probe current of 30 nA for metallic phases and 15 kV and 20 nA for oxides. Focused beam with 1 μm diameter was used for metallic phases, 3 μm for oxides, counting time of 20 s on peak and 10 s on both (+) and (–) backgrounds were applied. The following standards, lines and crystals were used for metallic phases: pyrite (FeKα, LIF), sphalerite (SKα, PETJ; ZnKα, LIF), GeS (GeKα, LIF), AgMet (AgLα, PETH), chalcopyrite (CuKα, LIFH), GaAs (AsLα, TAPH; GaLα TAPH), In2Se3 (InLα, PETH), CdS (CdLα, PETH), SnS (SnLα, PETL), galena (PbMα, PETL), Sb2Se3 (SbLα, PETL; SeLα, TAPH) and NiMet (NiKα, LIFL). Whereas the following standards, lines and crystals were used for oxides: albite (SiKα, TAP), tugtupite (ClKα, PETJ), sanidyne (KKα, PETJ), cassiterite (SnLα, PETJ), NiO (NiKα, LIF), cuprite (CuKα, LIFH), Sb2Se3 (SbLα, PETH), willemite (ZnKα, LIFH), crocoite (PbMα, PETH), fayalite (FeKα, LIFL), anhydrite (SKα, PETL), In2Se3 (InLα, PETL), AgMet (AgLα, PETL), GaAs (AsLα, TAPH; GaLα TAPH), GeS (GeLα, TAPH). Data were corrected using ZAF procedure. Thanks to this, it was possible to analyze the distribution of elements in the micro-area, as well as to determine the content of metallic and oxide phases.

Fractions were also examined using a scanning electron microscope with an energy dispersive spectroscopy system (SEM-EDS) (SU-70, Hitachi), and point, linear and area analysis of the chemical composition.

Results and analyses

The visual apparition of Ge-In-D is shown in Fig. 1. It can be said that Ge-In-D are a heterogeneous mixture of metallic particles (inclusions, drops) and gray-brown fine, dusty powder. Occasionally, white pieces of ceramics could be found, which were the lining of the furnace.

Fig. 1
figure 1

Germanium-indium drosses; fractions of metallic inclusions and drops marked with circles.

Full size image

Figures 2, 3 and 4 show optical microscope photos of various grain classes of Ge-In-D. The material is heterogeneous in particle shape and morphology, and possibly composition, as evidenced by the different colors. In Fig. 4 can be clearly seen metallic copper (A, orange piece in the middle of the photo) and a blue-colored phase (B), probably Cu, Zn or Cd alloy. The structure of round grains is characteristic, which consist of a clear “shell” (in the photos as a dark outer shell of the grains) and a light “core” (Figs. 3B and 4: grains with a light and blue center). Segregation within grains may imply the chemical composition of individual parts and, consequently, different chemical composition results depending on the research method. Hence, there was a need to examine the chemical compositions of individual parts of the grains, especially the “shell” and “core” of the grains.

Point and linear analyzes of the chemical composition of grain classes show that they are not chemically homogeneous. Below is an example with a grain class of 1.0/2.0 mm (Fig. 5; Table 2). The “core” of the grains consists mainly of either a copper-tin alloy in the proportions 38.23/56.67 (wt%), lead with small admixtures of other metals (up to 82–92 wt% Pb) or from a Sn-Pb alloy in the proportions 15.75/82.37 (wt%). However, the “shell” consists of Zn − 39.12 wt%, Sn − 27.40 wt%, but most importantly, also in large amounts of Ge and In, 17.48 and 3.46 wt%, respectively. A linear analysis of the chemical composition along the grain (Fig. 6) shows that Ge rather accumulates in the “core” (up to 22 wt%), while indium is located primarily in the “shell”. The main component of the “core” is Cu (on average 66 wt%), and the “shell” consists of tin (up to 66% by weight) and Pb (up to 88 wt%), as well as some Cu.

Fig. 2
figure 2

Optical microscopy images of grain class 1.0/2.0 mm, 100x magnification.

Full size image
Fig. 3
figure 3

Optical microscopy images of grain class 0.160/0.20 mm, (A): 10x, (B): 20x magnification.

Full size image
Fig. 4
figure 4

Optical microscopy images of grain class < 0.040 mm, 50x magnification.

Full size image
Fig. 5
figure 5

BSE image of grain class 1.0/2.0 mm: (A)—points of point composition analysis, (B)—line of linear composition analysis.

Full size image
Table 2 Point composition analysis of grain class 1.0/2.0 mm using WDS-EDS method (as in Fig. 5A).
Full size table
Fig. 6
figure 6

Linear composition analysis of grain class 1.0/2.0 mm using WDS-EDS method (as in Fig. 5B).

Full size image

The results of point and linear analyzes for another grain class, 0.160/0.20 mm (Figs. 7 and 8; Table 3) show that the color of the area in the photos depends on the chemical composition of the examined area. The lightest areas (point 3, Fig. 7) are Pb (93.215 wt%) with a small Sn content (3.462 wt%). The slightly darker areas (points 6 and 7) consist of Sn (approx. 56 wt%), As (approx. 29 wt%) and Pb (approx. 11 wt%) and do not contain Ge. The gray areas (points 1 and 2) are a Cu-Sn alloy with an average composition of 64 wt%/35.5 wt%, Cu/Sn, respectively. Finally, the darkest areas (points 4 and 5) are an alloy with an average composition of 70–72 wt%. Cu and 27.5–30 wt% As. The investigated grain turned out to be low in Ge and In. However, the linear analysis of the composition (Fig. 8) shows that Ge and In are located primarily in the “shell” with concentrations of approximately 30 and 6 wt%, respectively; this also applies to Zn, which contained up to 63 wt%. The “core” includes Cu, Sn and Pb in various proportions.

Fig. 7
figure 7

BSE image of grain class 0.160/0.20 mm: (A)—points of point composition analysis, (B)—line of linear composition analysis.

Full size image
Table 3 Point composition analysis of grain class 0.160/0.20 mm using WDS-EDS method (Fig. 7A).
Full size table
Fig. 8
figure 8

Linear composition analysis of grain class 0.160/0.20 mm using WDS-EDS method (Fig. 7B).

Full size image

Point analysis of the grain class < 0.040 mm (Fig. 9) showed that the dark areas are exceptionally rich in Ge and In, up to 30 and 38.5 wt%, respectively (Table 4), and the darkest ones are rich in Ge and Zn (approx. 31 and 51 wt%, respectively). This applies not only to the “shell”, but also to whole grains, the BSE image of which is gray or dark gray. However, linear analysis along the grain (Fig. 10) confirms that the “shell” is exceptionally rich in In, up to 69 wt%. and that it also contains Ge and Zn (a few wt%) and a lot of Sn. The “core” consists mainly of Sn and Pb. Hence, it can be assumed that the most In should be found in the smallest fraction of Ge-In-D.

Fig. 9
figure 9

BSE image of grain class < 0.040 mm: (A)—points of point composition analysis, (B)—line of linear composition analysis.

Full size image
Table 4 Point composition analysis of grain class < 0.040 mm using WDS-EDS method (Fig. 9A).
Full size table
Fig. 10
figure 10

Linear composition analysis of grain class < 0.040 mm using WDS-EDS method (Fig. 9B).

Full size image

Table 5 presents the chemical compositions of grain classes determined by the WDS-EDS method. Fe in all grain classes was on average from approx. 2 to 6 wt%. The Cu content ranges from 7 to over 20 wt%, of which it was the lowest for the grain class 0.040/0.056 mm. The concentrations of Ga, similarly to Ge and In, were the highest in the smallest grain classes, i.e. from 0.056 to < 0.040 mm, and were respectively: approx. 0.035; approx. 15 and approx. 16 wt%. Apart from the 0.071/0.100 mm grain class, the Ge content in the remaining grain classes was in the range of 6–12 wt%; in the case of In, it was in the range of 5–6 wt%. As with a content of 12–16 wt%. it was most abundant in thick fractions and in the 0.071/0.100 mm fraction. The Ag content was the highest in coarse grains, from > 2.5 to < 0.50 mm, and amounted to 4–6 wt%, while in the remaining grain classes it was only a maximum of 1.5 wt%. Sn content in all grain classes was in the range of 20–35 wt%. It turned out that the most lead, 30–35 wt%, is found in grain classes 0.160/0.315; 0.100/0.160; 0.071/0.100 and 0.056/0.071 mm. It quite often happened that the analyzed points consisted of either only one element or several, as evidenced by the recorded maximum concentrations of the analyzed elements. They were probably pure metals or their low-melting alloys. The average composition of Ge-In-D determined by the WDS-EDS method is as follows (wt%): 3.872 Fe; 15.764 Cu; 9.274 Ge; 9.782 As; 2.617 Ag; 7.875 In; 27.195 Sn and 20.737 Pb.

Table 5 Chemical composition germanium-indium drosses determined by WDS-EDS method.
Full size table

Conclusions

Taking all of the above into account the following conclusions were made:

  1. 1.

    Among the various metallurgical by-products containing germanium and indium known in the world, germanium-indium drosses are characterized by a very rich chemical composition in terms of the content of Ge and In, amounting to up to 10 and 6% wt%,, respectively. This is a global phenomenon.

  2. 2.

    The grain size and chemical composition of Ge-In-D is complex: a mixture of metallic and oxidized particles, as well as low-melting alloys.

  3. 3.

    The main ingredients of Ge-In-D are Sn in the amount of approx. 27–36 wt%, Pb in the amount of approx. 19–29 wt%,, Zn - approx. 8–11 wt%,. and Cu - approx. 8–22 wt%,. However, the uniqueness of Ge-In-D results from the presence of Ge and In in large amounts, approximately 7–10 and 2–8% wt%, respectively.

  4. 4.

    Important components of Ge-In-D are also As, in the amount of approx. 1–10 wt%, and Fe, in the amount of approx. 3–4 wt%. Depending on the grain class, the chemical composition is not uniform, and the highest concentrations of Ge and In were recorded for the smallest grains. These are spherical grains, consisting of a “shell” (outer part) and a “grain” (inner part), and Ge and In accumulate in the “shell”.

  5. 5.

    Despite the fact that the concentrations of Ge and In in the smallest fractions, i.e. 0.040/0.056 and 0.00/0.040 mm, can be as high as approx. 14–15 and 15–16 wt%, respectively, all fractions of Ge-In-D are rich enough in Ge and In and the entire material as such is suitable for chemical processing, without the need for mechanical enrichment.