Sample Preparation and Subsampling

Sample preparation transforms collected material into a test sample from which an analytical portion can be taken. Subsampling reduces the quantity retained while aiming to preserve the composition relevant to the measurement. These operations connect field collection to analysis: a precise assay of an unsuitable split is still an unsuitable estimate of the original material.^[1]

A preparation protocol must therefore match mineral distribution, particle size, sample mass, analytes and the intended test. No single crushing size, pulverising specification or retained mass is appropriate for every deposit and analytical objective.

Define the Measurement Before Preparing the Sample

Preparation begins with an agreed measurement and reporting basis. Total elemental composition, a method-defined extractable fraction and a mineralogical description are different objectives. Some require extensive size reduction; others need intact textures or particular fractions. Destructive preparation should not inadvertently consume the evidence required for another planned test.

The field sample, laboratory sample, test sample and test portion describe different stages or roles in the measurement process; a physical sample may serve more than one role. In this article, an assay pulp is pulverised laboratory powder, not a processing slurry. Eurachem's sampling guide explicitly includes physical preparation and selection of the test portion within the measurement chain. The meaning of the final result depends on which of those materials it is intended to characterize.^[1]

CIM's exploration guidance recommends sufficient representative material for analysis and retained reference purposes, with preparation appropriate to the material and constituents sought. Its recommendations provide a planning framework rather than a universal preparation recipe.^[2]

Receipt, Identification and Material Condition

On receipt, the laboratory and project need to resolve identity, requested work and sample condition. A bag label alone does not demonstrate that its contents correspond to the intended interval. Submission records, container identifiers and any discrepancy in count or condition need reconciliation before those identifiers propagate into certificates and databases.^[2]

Relevant observations can include received mass, moisture, damaged packaging and unusual material behaviour. These observations help distinguish original material characteristics from losses or changes introduced later. If a sample arrives incomplete, the limitation cannot be erased by assigning the normal preparation code.

Drying and moisture reporting need explicit conventions. A dry-basis concentration uses a different denominator from an as-received concentration. Preparation conditions also need to preserve the property being measured; changes involving losses, contamination or chemical alteration are not harmless simply because a nominal drying or grinding step was completed.^[3]

The practical response is to state which material is dried, how moisture information is obtained and whether any separate test requires preservation. A certificate can then be interpreted against an identifiable sample state rather than an assumed one.

Coordinating Particle-Size and Mass Reduction

Size reduction and mass reduction solve different problems. Crushing and pulverising reduce fragment dimensions; splitting or division reduces the quantity retained. If a small split is taken while the valuable constituent remains concentrated in rare coarse fragments, later grinding cannot recover the fragments excluded at that split.^[3]

Conversely, repeatedly processing the entire sample is not automatically justified. It may consume archives, introduce additional opportunities for contamination or add work without improving the decision. The sequence should retain enough material at each particle size to control the variability relevant to the mineral distribution. Validation needs actual material types, including heterogeneous or difficult cases, rather than only a convenient uniform sample.

Splitting performance depends on how material is presented and collected. Segregation before division, incomplete recovery and preferential passage of particular sizes or densities can alter the retained fraction. A stated split ratio describes mass allocation; it does not prove compositional equivalence.^[3]

Mixing is useful only within the limits of the material and procedure. It does not establish that every subsequent scoop is representative, especially when a material resegregates during handling. The retained and rejected fractions should remain linked so that later checks can address the specific reduction step.

Contamination, Loss and Special Mineral Behaviour

Contamination may arise between samples or from the equipment and materials used during preparation. Loss can be selective rather than proportional: a small missing fraction may matter if it carries a different concentration. Preparation blanks, recorded cleaning arrangements and tests on duplicate splits address different parts of this problem. Passing barren wash material through equipment is a cleaning operation; it is not an analysed blank result unless material is appropriately collected and tested for that purpose.^[4]

Coarse, unevenly distributed gold illustrates why mineral behaviour matters. In the ALS procedure described, the entire retained coarse fraction is assayed, while separate test portions characterize the fine fraction. The combined result uses the masses of the screened fractions, not an unweighted mean of their assays. This provides an approach to the material actually screened; it is not a way to restore gold already lost during field collection or an earlier split.^[5]

A larger or more elaborate assay is therefore not an automatic remedy for every discrepant result. The review should identify where the relevant heterogeneity persists and whether the retained material still represents the original sample. Alternative preparation requires testing and a recorded change in method, not retrospective selection of whichever result appears most plausible.

What Preparation Controls Demonstrate

Particle-size checks test whether the preparation product meets the agreed sizing condition. ALS's preparation quality-control note describes these checks separately from coarse duplicates, which repeat splitting and pulverising after the initial crushing stage. A satisfactory sieve result does not demonstrate freedom from contamination or adequate distribution of rare valuable particles. Agreement between those duplicates does not independently validate shared field collection or crushing.^[4][5]

A useful control arrangement distinguishes material entering before mechanical preparation from material introduced only as a prepared powder. The former may reveal effects absent from the latter. Similarly, a second portion of one powder tests less of the process than an independently prepared split from coarse material. CIM's estimation guidance explicitly distinguishes field, crushed-material and pulp duplicates.^[6]

The acceptance criteria should be related to the specific preparation and analytical objective. A manufacturer's illustrated sizing targets or duplicate performance examples are not universal requirements. Unexpected results warrant investigation of the material, stage and operating conditions, including whether the check itself was correctly identified and handled.

Retained Material, Repreparation and Release

Retained core, coarse rejects and pulps provide different opportunities for subsequent work. Repeating analysis on a pulp cannot test whether its parent coarse split was representative. Repreparing a coarse reject can investigate later preparation, but not material excluded before that reject was created. Its identity, preservation and remaining quantity must also be checked; retention alone does not demonstrate that archived material is unchanged or adequate for the new test.^[2][1]

Records should connect the original identifier with preparation dates, relevant masses, specified size targets, duplicate relationships, retained containers and analytical certificates. Changes, unusual losses and departures from the requested procedure need to remain visible. CIM's database guidance emphasizes keeping the method and original records and identifying the result selected when multiple assays exist.^[6]

Consider a hypothetical group of samples with acceptable powder sizing but inconsistent coarse duplicates. The immediate conclusion is not that the instrument needs recalibration. The disagreement could involve the coarse split, mineral distribution or subsequent independent preparation. Comparing control results at successive stages helps focus the investigation; it does not justify discarding every high value or averaging away a problem without explanation.

Preparation is complete only in the operational sense when the prescribed work is finished. Fitness for use additionally requires evidence that identity, composition and reporting basis remain suitable for the intended inference. That distinction allows the final assay to support geological interpretation without concealing limitations introduced before measurement.