Sample Preparation: The Critical First Step
Sample preparation quality determines the reliability of assay results regardless of the analytical method used, for gold exploration samples, the standard preparation sequence is:
- Coarse crush. jaw crusher to 70% passing 2 mm
- Split. riffle splitter or rotary splitter to obtain representative sub-sample (typically 250–500 g)
- Fine grind (pulverise). disc mill or LM2 pulveriser to 85–90% passing 75 µm
- Sub-sample for assay. 30–50 g aliquot weighed for fire assay
Sample preparation introduces more variability into assay results than any other step. Thorough equipment cleaning between samples (compressed air blow-out between samples, full wet-clean periodically) is essential to prevent cross-contamination, for nugget-prone samples (coarse visible gold) a screen fire assay or metallic screen fire assay is required.
Analytical Methods for Gold
Fire Assay with AAS or ICP-OES Finish
Fire assay (FA) is the industry-standard method for gold determination in exploration samples. A 30 g pulp aliquot is fused with a litharge flux, the resulting lead button is cupelled to a gold-silver dore bead, and the bead dissolved in dilute nitric acid for analysis by atomic absorption spectrometry (AAS) or ICP-OES. Detection limit is typically 0.005–0.01 g/t Au, for samples with visible gold or evidence of high nugget effect, metallic screen fire assay on a 500 g or larger aliquot is required.
Multi-Element ICP-MS and ICP-OES
Pathfinder element suites (As, Sb, Bi, Te, W, Mo, Cu, Pb, Zn, Ni, Co, Cr, Ti, Ba, Sr, La, Ce) are determined by ICP-MS or ICP-OES following aqua regia or four-acid digest. Multi-element packages (typically 33–51 elements) provide geochemical context for mineralisation interpretation and are essential for alteration studies and understanding deposit-style controls.
QA/QC Programme Design
Certified Reference Materials (CRMs)
CRMs are independently certified pulp standards with known gold and multi-element values. They must be inserted at a frequency of at least one per 20 samples (5% of the batch), spanning the low-grade, mid-grade and high-grade range of the mineralisation. CRM results must plot within ±2σ of the certified value; failures require investigation and re-assay of the affected batch.
Blanks
Coarse crush blanks (barren rock of similar lithology to the samples, crushed and prepared identically) detect contamination in the sample preparation circuit. A blank result exceeding 5× the detection limit triggers a cleaning investigation. Pulp blanks detect cross-contamination within the analytical batch at the laboratory.
Field and Pulp Duplicates
Field duplicates (quarter-core or split-RC pairs) measure sampling precision. the reproducibility of the sampling method. Pulp duplicates (portion of the original pulp sent to a second laboratory) measure analytical precision. Duplicate datasets are assessed by HARD (Half Absolute Relative Difference) statistics; precision better than 20% HARD at all grade ranges is the target standard.
Laboratory Selection in Tanzania
Several internationally accredited commercial laboratories serve the Tanzanian exploration market from Dar es Salaam, with sample receipt and results turnaround of 5–10 business days. Key accreditation to look for: ISO 17025 certification for gold and multi-element methods, SANAS or UKAS accreditation, and participation in inter-laboratory proficiency testing programmes (AMIS, RockLab). Bart Mining assists clients with laboratory selection, contract negotiation and result database management.