Search Documents

By Kari Moen, John Rasmus Leinum, Jarle Hjelen, Terje Malvik

"Industrial minerals may contain the same elements in approximately the same amount. In these cases backscatter electron imaging or chemical analysis are not sufficient to identify the minerals. Another problem arises when micro textures are not detectable in traditional imaging or the grain size is smaller than the X-ray spatial resolution.As a supplementary technique in these cases, application of crystallographic information by electron backscatter diffraction (EBSD) in the scanning electron microscope (SEM) may be a suitable method. The technique has been demonstrated on three Norwegian mineral raw materials and one Italian marble and shows that it has some advantages compared to traditional techniques.• quantitative measurements by ebsdCommon automatic techniques describing the mineral contents and the occurrence of minerals in free and locked particles, are based on optical-, backscattered electron (BSE)- or X-ray-signals. There are various commercial software systems on feature analysis by backscattered and X-ray signals in combination with image analysis (e.g. Oxford Inca Feature, CSIRO QemScan).Especially in silicates, the occurrence of similar elements in the different minerals is common. Sometimes the various minerals not only contain the same elements, but also approximately the same amount of each element. In these cases BSE imaging and chemical analysis are not always sufficient for identification."

Jan 1, 2003

By H. Tiefel, M. Schneider, T. Neesse, J. Regler, J. Dueck, F. Donhauser

In practice the advantages of high-quality hydrocyclone separation are often reduced due to varying feed suspension properties, solids concentrations and particle size distributions. Control systems of large hydrocyclones are usually fitted with controllable underflow nozzles, which is not practical for small hydrocyclone batteries. Therefore, a new control system for a battery of hydrocyclones with a nominal diameter of 150 mm was invented at the LUR in cooperation with AKW Apparate und Verfahren GmbH & Co. KG. This process was developed for bentonite slurry regeneration associated with tunnel excavation using bentonite slurry as a transport aid and lubricant.

Jan 1, 2003

By E. Lee

Cone crushers are widely used within the mining and aggregate industries for the size reduction of minerals, ores, and rock materials. In mining applications for example, the product from cone crushers is often fed to HPGRs or directly to mills in the subsequent grinding circuit. An improved cone crusher performance can therefore result in an overall improved and more efficient comminution process. The objective of this paper is to systematically study a number of different ways of applying compressive crushing. By mathematically modelling these different crushing concepts and then theoretically optimizing crushing sequences accordingly, each concept can be studied and evaluated. In addition, aspects concerning whether or not the results are mechanically or practically achievable are considered. This study thus explores the advantages and disadvantages of different crushing concepts. In this paper, four different theoretical crushing concepts are presented and studied. The results show that in terms of interparticle crushing, optimal crushing sequences with varying compression ratios can be replaced by easier implemented series of constant compressions without significant deterioration. It is also shown that that considerably better performance is obtained for crushing sequences with both breakage modes. The subsequent comparison between such concepts indicates that different approaches can achieve the same results, and that the different concepts are comparable. It is further concluded that interparticle crushing should be kept to a minimum if fines generation or energy consumption is to be minimized, and that each of the studied concepts presents different challenges for implementation. Keywords: compressive crushing, cone crusher, optimization, modeling, crushing concept

Sep 1, 2012

By A. Bastrzyk

Text The mining and refining of minerals generates large amounts of solid waste. The waste from cleaning operations constitutes up to 90% of the total production. Although some of them can be used in mining, civil enegineering, construction, agriculture, most of them must be disposed in dumps and slurry ponds, causing major economic and enviromental problems (Alonso, 2002; Alon-so, 1999). The refuse from minerals cleaning may have considerable valuable particles content, which can be used in industry. Thus, the recovery of these particles from minerals waste by using a cleaning method effective at fine particle size, may overcome these consequences. Surfactants are used in numerous applications including emulsification and solid surface modification. In the oil agglomeration of minerals, the ionic surfactants (both anionic and cationic) are used for the hydrophobization of mineral surface and the oil-water emulsion preparation. When the cationic surfactant is injected into the oil-water mixture, surfactant molecules adsorb at the oil-water interface, changing droplets? properties. The addition of anionic surfactant to the mineral suspension was found to modify the solid-liquid interface. In this paper, the condition and mechanism of selective separation of magnesite from quartz are presented. A satisfactory was achieved with the amount of anionic collector (sodium oleate) equal to 15.2 mg/gsolid and with the amount of cationic collector 27.9 mg/gsolid for cetyltrimethylammonium bromide (CTAB) and 16.9 mg/gsolid for dodecyloammonium hydrochloride (DDAHCl) at pH 8. At these conditions, the amount of agglomerated carbonates was about 12% for CTAB and 5% for DDAHCl.

Jan 1, 2014

By John Ralston, Stephen Grano, Yongjun Peng, Daniel Fornasiero

"A specially designed mill which allowed the control of pH throughout grinding was used to study the effect of rgrinding conditions on galena flotation and galena separation from pyrite. The various reactions occurring on the mineral surface were investigated by a range of techniques, including OH- addition to maintain the grinding pH, EDTA extraction, XPS and ToF-SIMS measurements.Galena flotation and galena separation from pyrite were strongly dependent on the metal oxidation species produced on galena and pyrite surfaces under different grinding conditions. Iron oxidation species depressed both galena and pyrite flotation, whilst lead oxidation species activated pyrite but had little effect on galena. Optimum galena flotation and galena selectivity against pyrite were achieved by selecting grinding conditions that enabled lead and iron oxidation production to be controlled."

Jan 1, 2003

By Artem Dolotov, Alexander Ter-Oranesyants, Saveliy Kaplan, Boris Axenov, Vasiliy Kovalev, Valeriy Tsyplakov, Konstantin Grigorjev, Nikolai Vorob’ev-Desyatovskii

The issues of double refractory gold ores pressure oxidation were considered. It was demonstrated that along with chloride autoclave preg-robbing which has been studied there are other negative external and intra-autoclave processes resulting in gold losses. It was found that an ingress of even low concentrations of NaCN, ferrum(II) and ferrum(III) cyanide complexes, as well as thiocyanate compounds into the autoclave leads to decrease of gold recovery during further cyanidation in a "carbon in leach" process (CIL). Admixtures depressing recovery of gold can get into the autoclave together with water in the result of industrial accidents or due to peculiarities of recirculation of industrial water used for feed pulp preparation or with quench-water used for process temperature maintenance. In contrast to a chloride preg-robbing which occurs only at pressure oxidation of double refractory gold ores, cyanide preg-robbing starts in the autoclave feeding tank (so-called «upper» preg-robbing). Possible causes of cyanide autoclave preg-robbing and requirements to the quality of water supplied to autoclaving were discussed.

Jan 1, 2014

By V. Grujic, M. Grujic, D. Salatic

"Two areas are involved in the research work of the comminution process (crushing and grinding). The first one involves the changes of the size class distribution values that are definded by the differential equations in the function of the ore property and working conditions. The second area involves the comminution of the different levels in order to optimize the working parameters and to decrease the operational costs in the grinding process.We have studied, in detail, the possibility of functioning the crusher at the maximal loading, over 4500 kN, ball mills with diameters larger than 5.0 m and semiautogenous mills with diameters larger than 10.0 m.Using the computers, we have achieved the maximal loading range of the crusher, over Fc ? 4500 kN in the crushing process. In the grinding process, using the adequate differential equations in the automation process, we have established the correlation between the pulp mass capacity in the mill, pulp dencity, ore fineness in the pulp and recirculating ore charge.With the appopriate text, equations and graphs, this paper shows the results of our several years research work in the laboratory and in the ore comminution plant of Majdanpek Copper Mine."

Jan 1, 2003

By Kirsten Corin, Margreth Tadie, Michael Nicol, Cyril O’Connor, Jenny Wiese

"The recovery of base metal sulfides has been emphasized in the study of the flotation of platinum group minerals (PGMs). However, in some ore bodies such as the Platreef of the Bushveld Complex, sulfide minerals do not make up the majority of the more significant PGM mineralization. It is therefore important to develop a fundamental understanding of how non-sulfide and sulfide PGMs respond respectively to the current reagents being used in the flotation of these minerals. The interactions between platinum and palladium sulfides and tellurides with the thiol collectors, ethyl xanthate and diethyl dithiophosphate, have been studied using the electrochemical techniques of mineral rest potential and voltammetry measurements. The differences in the interaction between the two types of minerals with these reagents were of key interest. Rest potential measurements indicated stronger electrochemical interactions of the minerals with ethyl xanthate than diethyl dithiophosphate. These measurements also indicated that there are different forms of oxidized collector present on the mineral surfaces after interaction with ethyl xanthates. Voltammetry measurements support these observations and suggest differences of adsorption on Pt and Pd sulfides than on Pt and Pd tellurides. Voltammetry also showed that there was minimal electrochemical interaction between the Pt and Pd minerals and diethyl dithiophosphate."

Jan 1, 2014

By Mohan Ram

Structural impurities in the crystal lattice of rock phosphate of Indian origin is a great challenge for beneficiation and practically possesses remote scope of upgradation to the desired level for industrial utilization. There is a wide occurrence of rock phosphate in different parts of India with different characteristics viz. siliceous, carbonaceous and ferruginous. The structural cationic replacement in the composition of apatite is due to the exchangeable cations present in the external impurities associated with the rock phosphate ore. The Indian rock phosphate ore from Purulia district, West Bengal is characterized as unusual structural composition due to the exchangeable cation in its composition. The elemental distribution of ore exhibits the disproportionate ratio of P2O5 and CaO which is 5.88: 3.31 in reference to the theoretical composition of P2O5: CaO in apatite which is 4.23: 5.55. The substitution by trivalent Al+++ cations within the structure of apatite in place of divalent Ca++ is found upto 18% and substitution of exchangeable cation in place of PO4 and F ion by CO3/SO4 is found upto 0.1% and OH ions upto 0.2% respectively in a lattice of apatite thereby diluting the grade of rock phosphate and restricting its industrial use even after beneficiation. This paper reveals the unusual compositional features of rock phosphate under study and is a guideline to the mineral industries using such rock phosphate. Keywords: external impurities, exchangeable cation substitution, structural compositional features, electron probe micro analyzer

Sep 1, 2012

By P. Harris

"Comparison of the two most commonly used classes of depressant in the South African platinum industry is complicated by the fact that these polymers induce different effects on the surfaces onto which they adsorb. The use of carboxymethyl cellulose (CMC), which is ionically charged in solution, can lead to increased particle dispersion in addition to the effect of reduced surface hydrophobicity (depression). Guar is typically only slightly charged in solution. Comparison of these two depressants is complicated since it is uncertain whether differences in their action is a result of either depression of dispersion effects.This work investigates the role of dispersion independently to that of depression in the batch flotation of a platinum bearing ore. An inorganic dispersant, sodium hexametaphosphate (NaHMP) was added in parallel to the uncharged guar depressant to simulate the action of depressants with varying degrees of dispersion. By distinguishing between the material recovered by entrainment and the floatable gangue and by including the effects of froth stability it was possible to decouple the effects of dispersion and depression.The results show that both dispersion and depression led to increased grade of the sulphide minerals. Depression leads to a decreased recovery of floatable gangue minerals. Dispersion leads to an increased recovery rate of these minerals and an accompanying decrease in froth stability."

Jan 1, 2003

By Robert Snow, Shibo Zheng, Wenyi Song, Xiaoqing Ma, Patrick Zhang

Numerous phosphate mineral depressants have been reported in the literature for use during anionic flotation of dolomite from carbonate fluorapatite in slightly acid circuits. In one part of this research, laboratory flotation experiments were conducted to compare the effectiveness of phosphate depressants to process ground, deslimed high-MgO Florida phosphate pebbles. Ten potential phosphate depressants were evaluated during laboratory anionic flotation of dolomite from phosphate in a slightly acid circuit using a sulfonated oleic acid soap plus oil as the dolomite collector. Eight of the depressants have been used by various investigators with foreign and domestic carbonate-phosphate ores. Four of the depressants were found to be reasonably effective during the current flotation study, namely: sodium tripolyphosphate (STPP), sodium hexametaphosphate (SHMP), tetrasodium pyrophosphate (TSPP) and diphosphonic acid (DPA). Without depressants, the sulfonate collector required to produce cell underflow phosphate products containing <0.80% MgO and MgO/ P2O5 ratio <.033 was 1.13 kg/ton of flotation feed at a flotation pH range = 5.5-6.0. Using this collector level with 0.91 - 1.36 kg/ton of feed of STPP, SHMP or TSPP yielded phosphate products analyzing 0.73-0.92% MgO and MgO/ P2O5 ratios = .027-.034 at 90.0-96.8% P2O5 recovery. Similar grade phosphate products were obtained using DPA as the phosphate depressant at 84.1-87.5% P2O5 recovery. When no depressant was used, a similar grade phosphate product was obtained at only 65.9% P2O5 recovery. Another part of the research aimed at reducing MgO content in the final concentrate from Florida phosphate plants by conducting dolomite flotation on the final concentrate. Numerous phosphate depressants were tested for this purpose, including acidic, neutral and alkaline reagents. The most effective phosphate depressant was found to be phosphoric acid, giving both high grade and recovery.

Jan 1, 2014

By C. H. Veloso

"Starch is the most widely used depressant in the reverse cationic flotation of iron ores. There is however, an indication from recent studies, that other polysaccharides may successfully replace them. When starches are used in the flotation of iron ores in the presence of quartz as gangue mineral, their strong adsorption on the surface of iron oxides can be explained by the molecular conformation between the available sites on the mineral surface and two hydroxyls of starches. However, when the gangue minerals are iron-bearing silicates, the strong interaction with starch can occur on these minerals since they have the same surface behaviour as iron oxide. In this work bench scale tests were conducted in order to generate scientific knowledge about other alternatives depressants for the flotation of iron ores. A magnetic pre-concentrate from a skarn iron ore processing plant was used as feed material for flotation tests. The reverse flotation approach was used to upgrade the iron content by removing complex ironbearing silicates. The depressants used in this study were: corn starch, dextrin, carboxymethyl cellulose (CMC) and humic acid. The tests were performed under different pH conditions in order to try to relate the efficiency of depression to the pH used during flotation. The results show the important role of pH as each depressant reached its best performance at different pH values. At alkaline pH the corn starch showed best depression performance while the humic acid is more efficient at pH close to neutrality.INTRODUCTIONIron corresponds to approximately 6% of earth’s crust composition (McDonough and Sun, 1995). It is the most used metal in our lives, some of its properties as resistance and facility to be shaped makes is replacement very difficult. The primary sources of iron on our planet are minerals. The iron-bearing minerals most commonly found in nature and economically exploited are magnetite [Fe3O4], hematite [Fe2O3], goethite [FeO(OH)] and siderite [FeCO3] (L O Filippov et al., 2014). These minerals are commonly associated with other minerals (most of the time not economically exploited) forming the so-called iron ore. Iron ore is the source of primary iron for the world’s iron and steel industries."

Jan 1, 2018

By A. Otsuki

This paper describes the evaluation method of dispersion and coagulation of silica nano-particles by means of interactive force apparatus. The apparatus measures the size of fine particles in aqueous solution or organic solvent of high concentration and/or low optical transparency. Other conventional apparatus for size measurements (such as, apparatus utilizing dynamic light scattering) are difficult to measure the size of particles in such conditions accurately, due to multi-scattering of light in the sample solution or solvent. The apparatus measures the interactive force between two surfaces, i.e. the gold-coated glass hemisphere and the brass flat plate, (which is fixed at the bottom of the sample cell). The apparatus has three parts, i.e. main part (which consists of electric balance, hemisphere and flat plate), control part (i.e. personal computer, piezo-stage controller and voltage supplier), and detecting part (i.e. multi-meter and oscilloscope). During the measurement, electric field is applied between the hemisphere and flat plate, and thus dielectric particles are arranged toward the direction of electric field in the area between two plates. When the two plates are close to each other, two different forces (i.e. repulsive force and attractive force) will alternately act on particles due to the structure change of the particles. The repulsive force occurs when the structure of particles is stretched by compressive force; whereas the attractive force occurs when the aggregate structure is broken. The cycle of the repulsive and attractive forces is a primary size of particle or size of aggregate, which depends on the degree of agglomeration. The measured size of particles, which is varied depending on their degree of dispersion and coagulation, determined the characteristics of the silica particles. In this research work, the effect of size and concentration of particles in aqueous solution on the degree of dispersion and coagulation of the silica particles were investigated. Spherical silica particles were selected for this investigation due to their variety of size. This apparatus can be used to evaluate the degree of dispersion and coagulation of other nano-particles in various solvents in many process industries.

Jan 1, 2014

By Ayumu Tokumaru, Shuji Owada, Gangfeng Liu

"Several million spent catalytic converters are disposed annually worldwideandeach containing about 0.5-1g ofplatinum and 1-2g of all the platinum group metals (PGMs), then,it is fairly significant to recover PGMs from the spent automobile catalyst. The PGMs’ recovery is a proven technology, including pyro-and hydrometallurgicalmethods, and the highestrecovery can achieve approx. 98 wt.% by acid-leaching method, but theserecyclingtechnology seems to be too highlyenergy-consuming. SincePGMs aredistributed only in coat layer in spent catalyst, it is of great significance to enrich the coat layerwith some physical methods as a pretreatment.In this paper, according to the double-deck structureof the catalyst andthe different properties of the two main compositions, alumina-ceria phase of coat layer and cordieritephase of substrate, the concentration of the coat layer was investigated by combiningthree kinds of physical separation technologies, heating-quenching, crushing and surface grinding. In the first stage, thermal stresscouldgeneratebetween the two phasesby thermal expansion difference,to producefracturesin the thinnest part of coat layerand micro-cracksin coat layerand at the interface of coat layerand cordierite substrate. The second stage of crushing, by jaw and roll crushers, could break dominantly the coat layer and the third stage of surface grinding, carried out by attritor, promoted further separation and concentration of coat layer fromthe substrate.We also improved the attritor grinding by adding a new lift-arm to the cell for achieving better mobility of sample particles in the selective surface grinding."

Jan 1, 2014

By Anuttam Patra, Romain Bordes, Andreas Fredriksson, Tommy Karlkvist, Krister Holmberg, Kota Hanumantha Rao

The present investigation aims to develop and distinguish mineral specific reagents with two functional groups for use in flotation of calcium containing minerals. For this purpose, a series of dicarboxylate-based surfactants with varying length between the carboxylate groups (one, two or three methylene groups) were synthesized. As reference, a surfactant with the same alkyl chain length but with only one carboxylate group in the polar part was synthesized. The adsorption behavior of these new reagents on pure apatite and calcite mineral surfaces was studied using Hallimond tube flotation, FTIR and ? potential measurements. The relation between the adsorption behavior of a given surfactant on a specific mineral surface and its molecular structure over a range of concentration and pH values, as well as the region of maximum recovery were established. It was found that one of the reagents, with a specific distance between the carboxylate groups, was much more selective for a particular mineral surface than the other homologues synthesized. This selective adsorption of a given surfactant to a particular mineral surface relative to other mineral surfaces as evidenced in flotation studies is substantiated by ? potential and infra-red spectroscopy data.

Jan 1, 2014

By Leon Lorenzen, Rodrick M. Lamya

"Leaching behaviour of nickel – copper matte in a CuSO4–H2SO4 solution during the repulping stage of a matte pressure leaching process has been studied. The base metal recovery process investigated in this study consists of a matte repulping stage prior to a series of pressure leach stages. The repulping stage is essentially an atmospheric leach stage in which the Ni-Cu matte, containing small amounts of cobalt and iron, is leached in copper spent electrolyte from the tankhouse, resulting in dissolution of the nickel with simultaneous precipitation of almost all the copper.Effects of various process parameters on the kinetics and extent of the leaching reactions, as well as on the cementation of copper from the solution have been quantified. The parameters investigated were temperature, particle size, pulp density, stirring speed, residence time, initial copper concentration and initial acid concentration.Leaching characteristics of the preleached matte during the subsequent pressure leach operation has also been studied. The effects of variations in the conditions in the repulping stage on the kinetics and degree of base metal dissolution during the pressure leach stage have been investigated. Any transformations that occur in the mineral phases of the matte during the leaching process have been studied, and a leaching mechanism proposed."

Jan 1, 2003

By Kumar Ashish, Shukla Akhilesh, Kr. Rokkam Kiran, M. Sheeba, S. Sombhatla Sundar

"Purification of zinc leachate plays a predominate role in zinc leaching refineries, to produce impurity free cathode sheets. As a part of such purification sequences, copper cementation with zinc dust is the preliminary step, since the presence of copper will preferentially deposit during zinc electro deposition and has the effect of reducing current efficiency. The cake thus produced is being accumulated as an inventory, due to invariable market and leading to space constraints & hazardous ambience. Hence, a project has been taken up for in-house generation of copper sulphate solution through ammonical leach followed by solvent extraction process and utilizing it as an active promoter for zinc sulphide flotation in Hindustan Zinc’s own captive mines. Cold purification cake majorly contains copper (30-40%) along with zinc (10-20%) and cadmium (2%). Confirmation to the above statement, photomicrographs of purification cake also confirmed the major presence of copper in metallic or oxidized form. In present work ammoniacal carbonate leaching of purification cake followed by solvent extraction using diketone based solvent has been studied under the influence of various parameters viz., temperature, agitation, pulp density, ammonia, CO2 dosages, solvent concentration, and impact of w/s zinc on leaching and solvent extraction. Leaching kinetics determined based on shrinking core model. Chemical reaction at unreacted core was found to be the rate controlling step. The estimated activation energy was found to be 24 KJ/mol. Leached copper has been extracted by solvent extraction with a ß diketone based solvent and is stripped with sulfuric acid as concentrated copper sulfate solution. The above established R&D findings are successfully implemented in the Commercial plant with a treatment capacity of two tons of cold purification cake per batch. Till now 500 MT of Purification cake has been treated, and generated 6500 MT of copper sulphate (55 g/l Cu), with final Cu recoveries between 87-90%."

Jan 1, 2018

By X. Y. Zeng

Most Chinese titanium bearing minerals are the composition of magnetite vanadium and ilmenite. All of such mineral processing plants use low intensity magnetic separators to recover magnetite and vanadium goes with magnetite. In the past years most of the plant tails containing about 10% TiO2 were abandoned. Now many of these plants use SLon magnetic separators to recover ilmenite in the roughing stage, which can throw away most of the gangs and slimes with very low cost. The new development of applications of SLon magnetic separators in Chinese ilmenite processing industry is introduced in the paper.

Jan 1, 2014

By Marco Le Roux, Martha Van Rensburg, Quentin Campbell

Fluidised bed drying is currently receiving much attention as a dewatering option during the beneficiation of fine (defined here as between 1 mm and 2 mm particles) coal. Apart from concerns about safety and combustion, the operating costs can be high if very high gas or air temperatures are used. The aim of this study was to investigate the removal of moisture from fine coal by using air at a lower temperature (25 - 55°C) within a controlled environment by lowering of the relative humidity of the air. It was firstly proven that fluidization has a major advantage over non-fluidized beds, with regards to drying rate and final moisture content. Hereafter several parameters that influence the dewatering rate and final moisture content were tested, from which it was concluded that the relative humidity of air acts as the largest driving force for dewatering. Parameters tested included relative humidity, temperature and coal composition. Energy balances showed a positive nett energy increase in the dried coal during combustion when related to the energy input during drying. In comparison with thermal drying technologies, fluidized bed drying was proven to be less energy intensive. It was therefore shown that this is a viable technology for the dewatering of fine coal.

Jan 1, 2014

By I. Y. Titov, Victor D. Samyguin, A. Y. Nikitin

"It is suggested that improvement of separation process efficiency in machines and circuits of selective flotation should be based on application of the principle of technology individualization (PIT). This principle is grounded on the following provisions.The mineral composition of ores is used to establish the correlation between the selective reagent mode and the individual value of hydrodynamic conditions of flotation machines together with individual structure of the process flow diagram. The individuality of all sorts of particles identified by methods of process mineralogy was characterized by their distribution according to the rate of flotation (or spectrum). The spectrum identification is carried out by modeling and by experimental and computational method on the data obtained on various types of laboratory flotation machines.The computer-aided test is used to determine the optimal fraction distribution, which reaches the highest possible level of concentration data. The direction of hydrodynamic factors action on concentration data is estimated based on comparison of design base and experimental spectrum. The improvement of separation efficiency in devices is achieved in special-purpose flotation machines (SFM) the design of which provides optimization of hydrodynamic regime in spatially isolated zones. It is demonstrated that the main parameter of SFM operating mode individualization is the rise rate of bubbles going from the chamber into the froth. During this time the composition of particles optimum for separation is formed on its surface. Additionally, it is necessary to coordinate in the zones the residence time of pulp and foam streams with the air consumption. The improvement of separation efficiency in flotation process flow diagrams may be achieved by making several cascades with SFM installed in them. The individual effective structure of the scheme for a definite ores material composition depends on the ratio of particle varieties in the spectrum and is revealed by modeling."

Jan 1, 2018