← Mineral Exploration 101
Sampling & Assay Practitioner / Student

How Assays Work: Fire Assay, ICP, XRF & Screen Fire

An assay is a laboratory measurement of the concentration of a target metal in a rock sample. Fire assay, ICP, XRF, and screen fire assay are the four methods you will encounter most often in mineral exploration — here is how each works, when to use it, and what its limitations are.


An assay is a laboratory measurement of the metal content in a rock, soil, or sediment sample, expressed as a concentration — grams per tonne (g/t) for gold and silver, or percent (%) for base metals such as copper, zinc, and lead. Assay data is the quantitative foundation of every resource estimate, drill intercept announcement, and mine plan. Without it, you have descriptions of rock — with it, you have numbers you can model, report, and ultimately mine.

This guide covers the four analytical methods you will encounter in the vast majority of exploration programs: fire assay, ICP (inductively coupled plasma), XRF (X-ray fluorescence), and screen fire assay. It also explains how samples are prepared before any analysis takes place — a step that is just as important as the measurement itself.

If you have not yet read Core Sampling & Splitting: Half-Core, Quartering & Why It Matters, start there — the assay is only as good as the sample that arrives at the laboratory.


What happens to your sample before it reaches the analyser?

Before any chemical or physical measurement takes place, the rock sample must be converted from a chunk of core or sediment into a fine, homogeneous powder. This preparation stage is where many assay errors originate, and it is critical to understand.

The typical preparation sequence for a hard rock drill core sample is:

  1. Drying — the sample is dried at approximately 105°C to remove surface moisture, which would otherwise affect the recorded weight and any moisture-sensitive digestion steps.
  2. Crushing — a jaw crusher reduces the sample to fragments of roughly 10 mm or less. The full sample is crushed, not a sub-split, because gold and other metals can be unevenly distributed and reducing the particle size before splitting improves representativeness.
  3. Splitting — after crushing, the sample is split to reduce the mass while retaining representativeness. A riffle splitter or rotary splitter is used; a standard preparation protocol might retain 250–500 g for pulverising and reject the remainder.
  4. Pulverising — the split portion is pulverised in a ring mill (LM5 or equivalent) to a fine powder, typically 85–95% passing 75 microns (0.075 mm). At this particle size, metal is distributed as uniformly as it will be across the sample mass.
  5. Sub-sampling — a 30–50 g aliquot is taken from the pulp for the fire assay or digestion, or a larger portion for screen fire assay.

The laboratory cleans its equipment — crusher, splitter, and ring mill — between samples using a coarse silica flush or a dedicated sand blank to prevent cross-contamination. Failures at this stage are one of the most common sources of apparent contamination in a QAQC review.


Fire assay: the benchmark for gold

Fire assay is the accepted standard method for determining gold concentrations in rock samples. It has been practiced for centuries and has been the industry benchmark for exploration since the mid-twentieth century. Every significant gold resource estimate in the world is built on fire assay data.

How fire assay works

The process has four stages:

1. Fusion. A 30 g or 50 g sub-sample of the pulverised rock is mixed with a flux — a blend of lead oxide (PbO), silica (SiO₂), borax, soda ash, and other reagents adjusted to the sample’s geological matrix. The mixture is placed in a clay crucible and heated in a furnace at temperatures above 1,000°C. At this temperature the rock’s crystal lattice is completely broken down and the gold (along with silver and platinum group metals) is collected by the molten lead into a heavy lead button that sinks to the bottom of the crucible, while the gangue material forms a slag above it.

2. Cupellation. The solidified lead button is removed and placed in a porous bone-ash cupel, then heated again in an oxidising furnace at around 950°C. Under these conditions the lead oxidises to lead oxide (litharge), which is absorbed into the cupel walls. The process is complete when the small shining bead of precious metal — called the doré bead or prill — is all that remains.

3. Dissolution. The doré bead is dissolved in aqua regia (a mixture of hydrochloric and nitric acid), bringing the gold into solution.

4. Measurement. The gold concentration in solution is measured by one of three finish methods depending on the grade range expected:

Grade rangeFinish methodDetection limit (Au)
Trace exploration (< 10 g/t)ICP-AES or ICP-MS0.001 ppm (1 ppb)
Trace to mid-gradeAtomic absorption spectroscopy (AAS)0.005 ppm
High grade or ore controlGravimetric (weighing the prill directly)0.05 ppm

The ICP-AES and ICP-MS finishes give the lowest detection limits and are the standard for exploration work, where you need to detect subtle gold anomalies in background rock.

Why fire assay is trusted

The key advantage of fire assay is total dissolution — the fusion completely breaks down all mineral phases including refractory minerals such as arsenopyrite, pyrite, and tellurides that resist acid digestion. These minerals commonly host gold in orogenic and epithermal systems. An acid digestion on the same sample might dissolve the easily accessible gold but leave significant gold locked in sulphide lattices undetected; fire assay captures all of it.

The fire assay process also removes the entire geological matrix before measurement, eliminating spectral interferences that can affect direct digestion-measurement methods. This matrix removal is why fire assay results are consistently more accurate across diverse geological types than most alternatives.

What fire assay does not measure

Fire assay is designed for gold and silver, and by extension platinum group elements. It is not used for base metals (copper, zinc, lead, molybdenum) or trace multi-element suites — those require ICP methods with acid digestion. Most programs run fire assay for gold alongside a multi-element ICP package for the same samples.


ICP: multi-element analysis for base metals and pathfinders

ICP (inductively coupled plasma) is a family of analytical techniques that measures the concentration of elements in a dissolved solution by exposing the solution to a plasma torch at approximately 6,000–10,000°C. At these temperatures, atoms in the solution are excited and emit light at element-specific wavelengths (in ICP-OES, or optical emission spectrometry) or are ionised and measured by mass (in ICP-MS, or mass spectrometry).

The key advantage of ICP is breadth: a single ICP run can simultaneously determine 30–50 or more elements across a very wide concentration range — from major elements like iron and aluminium at percent levels down to trace elements like gold, molybdenum, and rare earth elements at parts-per-billion levels.

Digestion methods used before ICP

The geological sample must be fully dissolved before it enters the ICP instrument. The digestion method chosen determines which minerals are dissolved and, therefore, which elements are fully captured:

Digestion typeReagentsWhat it dissolvesCommon use
Aqua regiaHCl + HNO₃Most sulphides, oxides, carbonates; not silicatesTrace gold in soils, regolith; base metal exploration
Four-acid (total)HF + HCl + HNO₃ + HClO₄All minerals including silicates and refractory phasesWhole-rock geochemistry, complete multi-element suites
Sodium peroxide fusionNa₂O₂ flux + fusionTotal dissolution including chromite and cassiteriteSpecialist refractory minerals

For most base metal exploration programs, aqua regia digestion followed by ICP-OES or ICP-MS is standard. For whole-rock lithological characterisation, a four-acid digest is used because it dissolves the silicate matrix.

ICP-OES vs ICP-MS: what is the difference?

ICP-OES (sometimes called ICP-AES) measures the light emitted by excited atoms. It provides accurate results for major and minor elements (typically above 0.1 ppm) at high throughput. ICP-MS ionises the atoms and separates them by mass-to-charge ratio; it is significantly more sensitive, reaching detection limits in the parts-per-trillion range for many elements. For exploration work, ICP-MS is preferred where trace-level detection is needed — pathfinder elements like tellurium, bismuth, arsenic, and thallium that may indicate gold mineralisation nearby.

When is ICP used in exploration?

  • Multi-element geochemistry on soil, rock chip, and drill core samples to define and vector anomalies
  • Base metal (Cu, Pb, Zn, Mo, Ni, Co) assaying throughout a drill program
  • Pathfinder element detection (As, Sb, Te, Bi, Hg) for gold systems
  • Whole-rock geochemistry for geological characterisation
  • As a finish instrument after fire assay fusion (Au-ICP methods) for the lowest gold detection limits

XRF: fast screening, not the primary assay

X-ray fluorescence (XRF) works by bombarding a sample with X-rays, which cause atoms in the sample to emit fluorescent X-rays at energies characteristic of each element. Measuring the intensity of these fluorescent emissions gives element concentrations.

XRF comes in two configurations that serve very different purposes:

Laboratory XRF (pressed powder or fused bead)

In a laboratory setting, pulverised sample is pressed into a disc or fused into a glass bead and placed in a benchtop XRF spectrometer. Laboratory XRF provides accurate, precise results for major elements (Si, Al, Fe, Mg, Ca, Na, K, Ti, Mn, P) and many minor elements, and is widely used for whole-rock major oxide analysis. It is non-destructive — the sample is not consumed.

Limitations: Laboratory XRF has higher detection limits than ICP for trace elements and gold. It is not the standard for gold assaying in exploration programs.

Portable XRF (pXRF)

Handheld pXRF instruments have become ubiquitous on exploration projects over the last fifteen years. A geologist can scan a drill core, rock chip, or soil sample and get element readings in seconds, without sending anything to a laboratory.

What pXRF is good for:

  • Rapid field screening to identify mineralised zones and prioritise samples for laboratory submission
  • Grade control screening at mine sites where speed outweighs precision
  • Identifying lithological boundaries by major element ratios (e.g. Cu, Fe, S in the field)
  • Checking whether a sample batch submitted to the lab arrived in the right order (correlation check)

What pXRF is not suitable for:

  • Primary assay results for resource reporting. pXRF results cannot be used in a JORC or NI 43-101 compliant resource estimate as the primary analytical dataset.
  • Low-grade or trace-level exploration, where detection limits are insufficient. Elements at concentrations below roughly 10–50 ppm are generally unreliable on handheld instruments.
  • Samples with high mineralogical heterogeneity. The pXRF beam analyses only the surface area it contacts (typically 1–3 cm²), making it sensitive to surface texture, moisture, and the position of mineral grains directly beneath the window.

The correct role for pXRF in an exploration program is as a screening tool that guides decisions in the field — not as a substitute for laboratory assay.


Screen fire assay: solving the nugget effect in coarse gold systems

Standard fire assay performs excellently on disseminated gold deposits where gold is distributed uniformly at the micron scale. But in high-nugget systems — bonanza epithermal veins, orogenic gold deposits, placer targets — gold occurs as discrete coarse particles that are unevenly distributed within any given sample. In these environments, a 50 g fire assay aliquot drawn from a 250 g pulp may or may not capture one of those particles. The result is extreme variability between duplicate samples — values that swing by factors of 5–20× — that reflects sampling variance rather than true geological grade variation.

Screen fire assay (also called metallic screening or metallic screen check) was developed specifically to address this problem.

How screen fire assay works

  1. Larger charge. The entire sample — typically 1 kg of crushed material — is used rather than a 30–50 g aliquot.
  2. Wet screening. The full charge is screened through a fine mesh, typically 75 microns or 106 microns. The fine fraction (undersize, which is the bulk of the mass) passes through the screen; the coarse fraction (oversize) is retained.
  3. Weighing both fractions. Both the fine and coarse fractions are weighed to allow a mass-weighted average to be calculated.
  4. Separate assays: The coarse fraction, which contains all the coarse gold particles, is fire assayed in its entirety — it is not sub-sampled. The fine fraction is sub-sampled and fire assayed in duplicate (two aliquots) to produce an average fine-fraction grade.
  5. Composite result. The final reported grade is a weighted average: [(mass coarse × assay coarse) + (mass fine × avg assay fine)] ÷ total mass.

The logic is that coarse gold particles, once captured in the oversize fraction rather than distributed across a small aliquot, contribute accurately to the final result — they are no longer a random variable.

When is screen fire assay triggered?

Screen fire assay is not used on every sample in a program — it is substantially more expensive and time-consuming than standard fire assay. It is triggered when:

  • Visible coarse gold is observed in the core at the logging or sampling stage
  • Standard fire assay returns highly variable results between field duplicates in mineralised zones
  • The deposit type is known to carry a high nugget effect (coarse-gold veins, nugget-bearing laterites)
  • The geologist or QAQC review flags anomalously high variability in the assay dataset

The decision should be documented in the sampling protocol and the trigger criteria defined before drilling begins, not retrospectively.


How do these methods compare? A summary

MethodPrimary useElementsDetection limit (Au)Destroys sample?Cost (relative)
Fire assayGold, Ag, PGEsAu, Ag, Pt, Pd0.001 ppmYesModerate
ICP-OES/MS + acid digestBase metals, multi-element30–50+ elements0.001 ppm (with FA precon.)YesModerate
Lab XRFMajor oxides, whole rockMajor elements~1 ppmNoLow–moderate
pXRFField screeningMajor–minor elements~10–50 ppm (variable)NoLow
Screen fire assayCoarse/nuggety goldAu, Ag0.001 ppmYesHigh

What about PhotonAssay?

A newer technology worth knowing is PhotonAssay, developed by Australia’s CSIRO and commercialised by Chrysos Corporation. It uses high-energy X-rays to excite atomic nuclei and measure gold (and silver, copper) concentrations in as little as two minutes, without chemicals or sample consumption. PhotonAssay analyses a much larger sample volume than fire assay — typically the full 500 g–1 kg pulp — which gives it advantages in high-nugget systems without the added cost and complexity of screen fire assay.

PhotonAssay is available at a growing number of commercial laboratories including SGS and Intertek. Several gold explorers have begun using it as either a primary method or a rapid screening tool alongside fire assay. It is not yet universally accepted as a standalone primary assay method for JORC/NI 43-101 reporting, but adoption is accelerating and the regulatory position is evolving. It is worth monitoring if your programs involve coarse gold or high sample throughput requirements.


How does your assay result connect to what the geologist logged?

The assay return from the laboratory is a number — a grade against a sample number. For that number to be useful, it must be matched back to: the hole ID, the from-depth and to-depth of the interval, the sample number assigned in the field, and the geological log for that interval. Without all of these connections intact, the assay is an orphaned data point.

In practice this linkage — from the geological log to the sampling record to the laboratory dispatch form to the assay certificate — is where the most data management errors occur in exploration programs. Spreadsheet-based programs with separate tabs for logging, sampling, and assay data are especially vulnerable, because the manual matching process introduces mismatches that compound through resource estimation.


FAQ

What is the difference between an assay and a geochemical analysis? The terms are often used interchangeably in exploration, but there is a subtle distinction. “Assay” historically refers specifically to determining the content of a precious or valuable metal — gold, silver, copper — in a sample. “Geochemical analysis” more broadly covers the measurement of any element or suite of elements for geological or environmental interpretation. In practice, most exploration programs run both: a fire assay for gold and a multi-element ICP package (geochemical analysis) on the same pulp.

Why is fire assay the gold standard for gold, rather than something faster like ICP alone? ICP can measure gold in solution at very low detection limits, but the challenge is getting the gold fully into solution in the first place. Direct acid digestion of a rock sample (aqua regia, four-acid) may leave gold locked inside refractory sulphide minerals undetected. The fire assay fusion step uses extreme heat to break down every mineral phase, ensuring total gold recovery before ICP measurement. For gold specifically, the fire assay preconcentration step is what makes the measurement reliable.

When should I request a screen fire assay? Request screen fire assay when visible coarse gold is observed in core, when standard fire assay duplicates show high variability (coefficient of variation above ~30% in mineralised zones), or when the project’s geology document identifies a nuggety gold style. If in doubt, run screen fire assay on a subset of samples in the highest-grade zone first and compare the results to standard fire assay to determine whether the nugget effect is present and material.

Can I use pXRF results in a resource estimate? Not as primary analytical data. pXRF results cannot satisfy the sampling and analytical requirements of JORC (2012 Table 1) or NI 43-101 for a compliant mineral resource. pXRF data can be used as a co-variate or screening dataset alongside laboratory assay results, but all reported grade intervals and resource blocks must be based on accredited laboratory analyses.

What does “detection limit” mean for an assay, and why does it matter? The detection limit (also called the lower limit of detection or LDL) is the lowest concentration the method can reliably distinguish from zero. If a fire assay method has a detection limit of 0.001 ppm (1 ppb), samples with gold below this level are reported as < 0.001 ppm. Detection limits matter in exploration because many anomalous but sub-economic zones, and many regolith samples above mineralisation, contain gold concentrations that are only a few ppb above background. A method with a higher detection limit (say 0.01 ppm) will report many of these samples as zero, creating a data gap exactly where you need nuance to vector toward mineralisation.


How Blue Butterfly connects assay data to your geological record

Assay data is only useful if it arrives in the right place. Blue Butterfly imports assay certificates directly from the laboratory return file — CSV, Excel, or lab API — and links each result to the sampling interval already on file in the project database, matched by hole ID, from/to depth, and sample number.

There is no manual matching step, no VLOOKUP in a spreadsheet, and no risk of a transposed sample number turning a 5 g/t intercept into a data artefact. The complete chain from the geological log through the sampling record to the validated assay result is intact and queryable from the moment the file arrives. QAQC results — blanks, standards, and field duplicates — are automatically flagged against their certified values so analytical failures are visible before the data enters any interpretation.

See how Blue Butterfly handles assay data import and validation →


Sources

Built for Geologists

Core logging that respects your workflow.

Cloud-native, validation at entry, real-time sync. Set up by a geologist in an afternoon, not a consultant over a week.

  • 01 One live database, no version conflicts
  • 02 Built-in strip logs and 3D drill hole viewer
  • 03 Per-project pricing, not per-seat
Get Early Access

Free during beta · No credit card


← Back to Mineral Exploration 101