Dataset prepared for characterization of three South African manganese ores before or after preheating in laboratory-scale rotary kiln

Manganese ores are the major raw materials utilized in the production of manganese ferroalloys. A common problem in the production of manganese ferroalloys is the lack of knowledge regarding mineralogical and metallurgical properties of manganese ores. Decrepitation, which is described as the breakage or disintegration of the ore particles upon heating, is an important quality parameter of these ores. The decrepitation Index (DI), which is the parameter used to describe the extent of decrepitation, is described as the ratio of mass of particles <6 mm after pre-heating to the total mass of the sample. The purpose of this paper is to describe all the raw data produced during the course of an investigation into the effect of temperature, rotational speed, and particle size on the decrepitation of three types of manganese ores sourced from South Africa. Furthermore, the relevance of the raw data to scientific community as well as industry is explained. The data set will include sub-sets of data i.e. the characterization of the as-received samples, the decrepitation test results, and the characterization of post experiment samples.


a b s t r a c t
Manganese ores are the major raw materials utilized in the production of manganese ferroalloys. A common problem in the production of manganese ferroalloys is the lack of knowledge regarding mineralogical and metallurgical properties of manganese ores. Decrepitation, which is described as the breakage or disintegration of the ore particles upon heating, is an important quality parameter of these ores. The decrepitation Index (DI), which is the parameter used to describe the extent of decrepitation, is described as the ratio of mass of particles < 6 mm after pre-heating to the total mass of the sample. The purpose of this paper is to describe all the raw data produced during the course of an investigation into the effect of temperature, rotational speed, and particle size on the decrepitation of three types of manganese ores sourced from South Africa. Furthermore, the relevance of the raw data to scientific community as well as industry is explained. The data set will include sub-sets of data i.e. the characterization of the as-received samples, the decrepitation test results, and the characterization of post experiment samples.

Value of the Data
• The data will be useful when a production facility with pre-heater is constructed/ a preheater is retro-fitted in an existing facility. The production engineers/scientist will use the data to select ores or ore blends if necessary and kiln process parameters that will ensure the most possible optimum operation of the (SAF) Sub-merged Arc Furnace. Proper selection of ores/ ore blends will also help ensure safety of plant workers. • Utlilizing these data will help the management of the facility ensure the process runs smoothly with minimum unscheduled shutdowns, this will in turn lower production costs, maximize profits, and improve the economy of the country. • The data can also be used to further understand the behavior of minerals found in manganese ores when exposed to heat, this can give insights to scientist (mineralogist) who intends to design experiments to study the behavior of different minerals in manganese ores. Table 1 summarises the porosity, moisture content, and LOI expressed in percentage (measured in triplicate) of natural Ore A, Ore B, and Ore C as well as their bulk density at + 6-20mm, and DI at 800 °C and 6 rpm.  Table 2 summarises the calculated average (of three measurements) and standard deviations of the bulk chemical compositions of the natural Ore A, Ore B, and Ore C as determined by ICP-OES for three representative samples per ore. When analyzing manganese ores it is conventional to present Mn and Fe in elemental form and other components in oxide form due to the complex nature of the Mn and Fe bearing minerals in the ores. It also explains why the total compositions of the ores are significantly less than 10 0.0 0%.  Table 3 presents the bulk phase chemical compositions of the nartural Ore A, Ore B, and Ore C as determined by QXRD. Values marked with an asterisk ( * ) indicate that the absence of a specific mineral in that specific ore. Table 3 Minerals present in natural Ore A, Ore B, and Ore C and their quantities in mass percentage (wt%) as determined by QXRD. The ideal chemical composition of minerals present in ore samples are presented in Table 9 .

Mineral
Ore     To determine the bulk chemical and bulk phase chemical compositions of the samples after heating, each sample was sieved to prepare sub-samples containing the < 6 mm and > 6 mm size fractions. Table 4 to Table 6 . present the bulk chemical compositions of the < 6 mm and > 6 mm size fractions determined by ICP-OES on three sub-subsamples per sub-sample.
Only the > 6 mm sub-samples were characterized by QXRD to determine their bulk chemical compositions. The analysis was done for all three ores and the section marked with * show that a specific mineral in a specific ore was not detected.
The data under the "physical properties" tab in the spreadsheet on Mendeley data is similar to the data displayed in Table 1 . The only difference is that the online data shows data for individual analysis and averages while the data in Table 1 shows only averages of individual analysis. The same applies to the data in the "ICP-OES" tab, the data is similar to the data shown in Tables 4 , 5 and 6 . The data sets in the tabs "QXRD" and " DI charts" are the duplicates of the already shown in the manuscripts, ( Table 7 ). Table 4 Calculated average and standard deviations (Std Dev) of the bulk chemical compositions, determined by ICP-OES, of the < 6 mm and > 6 mm sub-samples of Ore A. The samples were heated for 30 minutes at 60 0, 80 0, or 10 0 0 °C in the rotary kiln with the rotational speed fixed at 6 rpm and the input material being the + 6-20mm size fraction. Values presented in mass percentage (wt%).   Table 7 Minerals present in heat-treated Ore A, Ore B, and Ore C and their quantities in mass percentage (wt%) as determined by QXRD. The + 6-20 mm size fraction was heat treated at 60 0, 80 0, or 10 0 0 °C at 6 rpm for 30 minutes in the rotary kiln.
Only the QXRD results for the > 6 mm sub-samples were presented here. The ideal chemical compositions of minerals present in ore samples are presented in Table 8 .

Methods
This investigation utilized three South African manganese ores form the Kalahari Manganese Field (KMF) named Ore A, Ore B, and Ore C on request of the supplier. In the order of 200 kg samples of each type of ore was sourced for the investigation. The samples were screened to obtain + 6-20 mm, + 20-40 mm, and + 40-75 mm size fractions using Endecotts screens. Each size fraction was coned and quartered to produce 10 kg samples representative of the size fraction in question. The sample is arranged in a cone shape on top of a plastic laid on the floor, a shovel is used to divide the sample into quarters. The 10 kg samples were subsequently split into 1 kg representative samples using a rotary splitter. The 1 kg samples were utilized in the decrepitation tests or for further characterization. The sample were taken for the following characterization techniques; Quantitative X-Ray Diffraction (QXRD), ICP-OES and SEM-EDS.
The samples were taken to determine moisture content, bulk density, porosity and loss on ignition (LOI). The moisture content was determined using the ASTM D2216-19 standard method, 1 kg of the ore sample was weighed and put in an oven to dry overnight at 105 °C [1] . After drying and cooling the sample was weighed again to determine the change in mass which was then calculated to be the moisture content of that specific ore. The loss on ignition (LOI) was determined using the standardized method from the ASTM D7348 standard [3] . 100 g of the sample was weighed, the sample was then placed in a muffle furnace and heated to 950 °C in an oxygen atmosphere. The sample was the cooled and weighed to determine the change in mass (mass lost due to ignition). The porosity was determined using ASTM D4872 method for He-pycnometry test [2] . A sample of particles ranging between 6 and 20 mm was weighed and added to the sample cup. The sample was then placed into a pycnometer chamber. Helium was added to the system and the system was left to reach equilibrium pressures. The system uses the equilibrium conditions to determine the porosity of the sample. A total of 3 samples were analysed for each ore and the average was calculated. The bulk density was determined using the modified volumetric method. The method involves filling a hopper of a known mass (with a volume of 4 293 cm 3 ) with particles of Mn ore and weighing it. The mass of the empty hopper was subtracted from the mass of the hopper and ore to get the mass of the ore. The method was repeated three times to ensure accuracy. The bulk density was calculated according to Eq. The bulk phase chemical analysis was determined by quantitative X-ray diffraction method (QXRD. The method provided an ability to quantify and characterize different crystalline phases as well as amorphous phases that may be present. A Bruker D8 diffractometer with an acceleration voltage of 35 kV and cobalt tube with Fe-low beta filter was used with 2 Ɵ angle ranging from 2 to 80 degrees and step size of 0.02 °2 Ɵ. The samples were then taken for quantitative analysis, the D-500 diffractometer with Cu K ά radiation and graphite monochromators was used. The refined parameters included scale factors, back-ground coefficients, peak width and profile parameters and cell dimensions. Values of occupancy factors were set to reflect the ideal stoichiometric compositions.
For the decrepitation tests which used a modified method from [4] , the laboratory-scale rotary kiln was switched on with a temperature set point of either 60 0, 80 0 or 10 0 0 °C, and a rotational speed of either 3, 6 or 12 rpm. Once the kiln was at temperature, 1 kg of ore of either the + 6-20 mm, + 20-40 mm or + 40-75 mm size fraction was fed into the kiln and maintained there for 30 minutes. After 30 minutes the sample was allowed to cool and screened for parti- Parameters marked with * are kept constant when investigating the effect of others.
cles < 6 mm. The DI was calculated according to Eq. (2 ). Where: • DI = Decrepitation index • M1 = Mass of particles < 6 mm in g • M2 = Total mass of the sample in g The experimental plan is tabulated in Table 9 . In one set of experiments, the temperatures were varied and the rotational speed and size fraction maintained at 6 rpm and + 6-20 mm respectively. For the next set of experiments, the rotational speeds were varied and the temperature and size fraction maintained at 800 °C and + 6-20 mm respectively. For the final set of experiments, the size fractions were varied and the temperature and rotational speed maintained at 800 °C and 6 rpm respectively.

CRediT Author Statement
Credit is given to the following individuals: • Dr Joalet Steenkamp for providing conceptualization, supervision, funding acquisition, resourcines and writing-review and editing • Professor Hillary Limo Rutto for providing conceptualization, supervision and writing-review and editing

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.