While processing nickel, typically sulfidic or lateritic ores are initially crushed and milled to a fine powder. Water is added for creation of a pumpable slurry that is treated with reagents during floatation for concentrating the nickel. The concentrate is further processed by smelting creating nickel rich matte while the impurities are removed as slag.
The matte that is rich in base metals is then converted, and leaching processes form a separate nickel stream. A copper stream is also obtained as sulfidic and lateritic ores are also rich in copper. Copper and nickel streams are purified in preparation for electro-winning in order to produce pure nickel and copper anodes. The resulting slag from smelting is processed for the recovery of base metals and PGMs. It is possible to use the slag composition for monitoring kiln health in the determination of tear-down time for the prevention of catastrophic failure. In this article, the analysis of filter cake, nickel concentrate and furnace slag is shown.
.jpg)
Figure 1.
Instrumentation
Model: |
Rigaku NEX CG |
X-ray tube: |
50 W Pd-anode |
Excitation: |
Indirect with polarization |
Detector: |
High performance SDD |
Analysis Time: |
600 sec |
Environment: |
Helium Purge |
Standard: |
15-position Sample Tray (32mm) |
Option: |
Manual Sample Press |
.jpg)
Figure 2.
Sample Preparation
The preparation of each sample is done by grinding to a dry, fine, homogenous powder of less than 200 mesh or less than 75 µm particle size using a ball mill or ring-and-puck shatterbox. In order for measurements to be done a sample is prepared by weighing 5g of the sample in a standard 32 mm XRF sample cup. The sample was then compacted with 500 inch-pounds of torque using a Manual Sample Press to ensure consistent compaction.
Calibration
A separate empirical calibration was built for each matrix type that includes concentrate, filter cake, matte and slag. Each calibration makes use of between 10 to 23 well assayed standards of the respective sample type, and alpha corrections are then employed to automatically compensate for variations in X-ray absorption and enhancement effects within the sample due to the independent variations in element concentration. A summary of each empirical calibration is shown here.
Table 1. Nickel Concentrate - 10 Standards
Component |
Concentration Range (mass%) |
RMS Deviation |
R2 Confidence |
MgO |
0.08 - 0.28 |
0.024 |
0.8983 |
Al2O3 |
0.09 - 0.24 |
0.023 |
0.9192 |
SiO2 |
0.09 - 0.98 |
0.034 |
0.9872 |
S |
31.00 - 33.30 |
0.496 |
0.7185 |
CaO |
0.10 - 0.71 |
0.032 |
0.9111 |
Fe |
34.20 - 38.20 |
0.75 |
0.8337 |
Co |
0.91 - 1.17 |
0.028 |
0.9454 |
Ni |
19.70 - 25.30 |
0.279 |
0.9899 |
Cu |
0.38 - 1.38 |
0.012 |
0.993 |
Repeatability
In order to demonstrate repeatability, one representative sample was selected from each calibration. The measurement of each sample was done in static position for 10 repeat analyses with a total analysis time of 600 sec per measurement, with typical results shown below.
Table 2.
Sample: Nickel Concentrate Units: % |
Component |
Standard Value |
Average Value |
Std Dev |
% Relative |
MgO |
0.14 |
0.148 |
0.003 |
2 |
Al2O3 |
0.16 |
0.152 |
0.002 |
1.3 |
SiO2 |
0.68 |
0.69 |
0.01 |
1.4 |
S |
31.8 |
31.84 |
0.03 |
0.1 |
CaO |
0.71 |
0.7 |
0.01 |
1.4 |
Fe |
36.9 |
36.2 |
0.06 |
0.2 |
Co |
1.02 |
0.99 |
0.03 |
3 |
Ni |
22 |
22.1 |
0.06 |
0.3 |
Cu |
0.44 |
0.42 |
0.003 |
0.7 |
Qualitative Analysis
Spectral analyses for each of the samples showed clean isolation of the respective peaks of interest. Select spectra from the RX9 polarization target and the Mo target show excellent peak separation of the main elements as well as the extremely low background produced by indirect excitation. The following legend is used: Red (Concentrate); Green (Filter Cake); Gray (Matte); Purple (Slag)
.jpg)
Figure 3.
.jpg)
Figure 4.
Discussion
The empirical approach shown here provides the most accurate results, with accuracy mainly determined by the goodness of the calibration standards and their assay values. Normally, suites of calibration standards are built from samples taken at the mine and processing sites and assayed using ICP. Samples that need to be used as calibration standards are selected to model each specific sample type by choosing samples that spread evenly across the expected concentration range for each respective element. This enables the use of matrix correction factors (also called alpha corrections) that model the expected variations in concentration of each specific matrix. When ore types are used that fall outside the ranges for a calibration, extra samples are accumulated and assayed. The additional standards just need to be added to the existing calibrations to define extended concentration ranges.
Conclusion
The Rigaku NEX CG combines indirect excitation with secondary targets, polarization targets, and a high-performance SDD detector to offer the right performance for elemental analysis of ores. During the entire processing cycle, especially during smelting, elemental and oxide composition of the ore material must be reliably monitored to ensure optimal process control and profitability. The Rigaku NEX CG analyzer is a suitable tool all through the quality control process. Furthermore, the NEX CG can be used for monitoring air filters for air quality emission control, and UltraCarry can be used to monitor process effluents to ppb range detection limits, making the NEX CG a versatile and valuable tool for several applications within ore mining and processing.

This information has been sourced, reviewed and adapted from materials provided by Rigaku Corporation.
For more information on this source, please visit Rigaku Corporation.