Minerals Hub / Innovation & Technology / How thorium gives a monazite deposit away from the air
Innovation & Technology · Monazite · 6 min read
How thorium gives a monazite deposit away from the air
A gamma-ray survey never sees a rare earth — it sees the thorium travelling with it, three decay steps removed, and every step is somewhere the inference can break.
Pending review

The short version
Rare earths emit no useful gamma signal of their own. What an airborne survey detects is thorium — and not even thorium directly, but a decay product, thallium-208, whose 2615 keV emission is the line the standard thorium window is built around. The reading is therefore a proxy for a proxy: it measures a daughter to infer a parent, and the parent to infer a mineral, and the mineral to infer a rare-earth grade. Each of those three steps holds most of the time, which is what makes the technique useful, and each fails in specific circumstances, which is what makes the map not a resource.
What the instrument is actually counting
An airborne gamma-ray spectrometer is a photon counter with energy discrimination: it sorts incoming gamma rays by energy and counts how many arrive in each of several defined bands. The bands are not all alike. Of the five standard windows for natural radioelement mapping, three are tied to a named isotope and a named emission line; the total-count and cosmic windows are tied to neither, and the cosmic window is stated with no upper bound.
The three radioelement windows are set out plainly in the guide to airborne gamma-ray survey specifications published by the Australian Geological Survey Organisation as Record 1995/60 — a 1995 document whose own framing is that its contents are "the minimum requirements". Potassium is read at 1460 keV, in a window running from 1370 to 1570 keV. Uranium is read at 1760 keV, window 1660 to 1860 keV. Thorium is read at 2615 keV, window 2410 to 2810 keV (AGSO Record 1995/60, Table 1.1).
Note what is not in that list. Neodymium is not there. Praseodymium is not there. No rare earth is, because none of them produces a gamma emission at an energy and intensity a survey aircraft two hundred metres away could distinguish from background. The elements a monazite deposit is valued for are, to this instrument, invisible.
The first inference: a daughter, not the parent
The thorium window runs from 2410 to 2810 keV, and the line it reads is 2615 keV — five kilo-electronvolts above the window's own midpoint, where the uranium window sits exactly on its line. And that photon is not emitted by thorium. It is emitted by thallium-208, one of the decay products whose activity is used to determine the activity of thorium-232.
This is not a limitation peculiar to airborne work; the same substitution is made in the laboratory. The 2023 assessment of monazite grades by Kotb and colleagues measured thorium-232 through its daughters — lead-212 at 238.6 keV, thallium-208 at 583.1 and 2614.7 keV, actinium-228 at 911.1 and 968.9 keV — because those are the emissions that exist to be counted.
Reading a daughter to quantify a parent only works if the chain is in secular equilibrium: every step disintegrating at the same rate as the one above it. The laboratory arranges this deliberately. Kotb and colleagues sealed their samples for four weeks precisely so that "daughter radionuclides undergoing disintegration at a rate equivalent to that of the parent radionuclides" could be assumed.
An outcrop has not been sealed for four weeks. Whether material in the ground departs from that sealed condition, by what mechanisms and how far, is not something the sources behind this article establish — so what the laboratory arranges deliberately is, in the field, an assumption the reading carries rather than a condition it verifies.
The second inference: thorium is not monazite
Suppose the equilibrium assumption holds and the survey has told you, correctly, how much thorium is in the top of the ground. You still have not been told what is carrying it.
Monazite is a strong candidate. As a general property of the mineral, it generally consists of approximately 55.0 to 60 per cent rare-earth metal oxides and 5.0 to 10.0 per cent thorium oxide (Kotb et al., 2023) — a statement about monazite as a mineral specimen, not about any deposit. Deposit figures often imply more, and the gap is a difference of basis rather than a disagreement. Where a project reports a rare-earth oxide grade and a monazite percentage for the same interval, that monazite percentage is a mass-balance estimate back-calculated from the assay rather than a direct measurement of the mineral; Osmond's published pairings for Orión imply a rare-earth content in the monazite above the general range quoted here, and the two cannot be reconciled from published material because the conversion factor is not published. None is computed here. A mineral that is between a twentieth and a tenth thorium by oxide weight is, from a hundred metres up, loud.
But it is not the only loud thing. Zircon carries thorium and uranium. Xenotime does. So do allanite, thorite, and a range of accessory phases that a heavy-mineral assemblage may contain in quantities too small to matter commercially and large enough to matter radiometrically. A thorium high tells you a thorium-bearing mineral is concentrated there. Which one is a question for a microscope.
The survey finds monazite's chaperone, not monazite. Everything after that is inference, and inference is where the cost lives.
The third inference: monazite is not grade
The last step is the one most easily over-read. Kotb and colleagues report "a positive correlation between the grade of the Monazite samples and the concentration of radioactivity" — a real relationship, and the reason the technique earns its place in a programme.
A positive correlation is not a calibration curve. The rare-earth content of monazite varies between deposits and within them, and the thorium content varies too, and they do not vary together in a fixed ratio. Two anomalies of identical intensity can sit above materially different rare-earth grades, because the thorium got there in a different proportion.
What the survey cannot fix
Two further constraints sit on top of the inferential chain, and both are physical.
The first is that gamma rays from the ground are attenuated by whatever lies between the rock and the detector, including water. The AGSO guide quantifies one case: an increase in soil moisture of 10 percent, from 10 to 20 percent, will decrease the airborne count rate in the thorium channel by approximately the same amount. Fly the same line in a wet season and a dry one and the map changes, without the geology changing at all.
The second is depth. Gamma radiation is absorbed within a short distance of rock and soil, so what a survey maps is the radioelement content of a thin skin at the surface. Mineralisation under enough cover is not a weak anomaly; it is no anomaly.
What the technique does, then, is narrow ground fast and cheaply, and it does that extremely well. It ranks targets. It traces the geometry of a palaeo-shoreline across country that drilling would take years to cover. It does not measure a rare earth, and no processing of the signal can make it do so — which is why the output of a radiometric survey is a place to put a drill rig, and never a number a reporting code will accept as evidence of grade.
Related
- Exploration (Mining & Production) — the programme these instruments serve
- Artificial Intelligence — the interpretation layer built on survey data
- Digital Mining — the systems that hold what the surveys produce
- Technology Providers — the firms that build and fly the instruments
Sources
- PRIMARYGrasty, R.L. and Minty, B.R.S., 'A Guide to the Technical Specifications for Airborne Gamma-Ray Surveys', Australian Geological Survey Organisation, Record 1995/60, Commonwealth of Australia, 1995; ISBN 0 642 22366 1, ISSN 1039-0073. NOT a Geoscience Australia publication — that name does not appear in the document, and 'GA7667' exists only in the URL filename. The document describes its own contents as 'the minimum requirements' (p. 5). Table 1.1, p. 6, 'Standard windows for natural radioelement mapping': potassium, 40K, 1460 keV, window 1370–1570; uranium, 214Bi, 1760 keV, window 1660–1860; thorium, 208Tl, 2615 keV, window 2410–2810; total count, no isotope, 410–2810; cosmic, no isotope, 3000 keV upward. Bi-214's three main emissions at 1730, 1765 and 1848 keV (p. 7). Soil-moisture attenuation of the thorium channel, stated for a single 10-to-20-percent step and explicitly approximate (p. 29).
- PRIMARYKotb, N.A., Abd El Ghany, M.S. and El-sayed, A.A., 'Radiological assessment of different monazite grades after mechanical separation from black sand', Scientific Reports 13, 15389 (2023). Published 16 September 2023. Composition, with the paper's own hedge: monazite 'generally consists of approximately 55.0–60% rare earth metal oxides, 24.0–29.0% phosphate (P2O5), 0.2–0.4% uranium oxide (U3O8), and 5.0–10.0% thorium oxide (ThO2)' — a general property of the mineral, not a measurement of these samples. Th-232 activity determined via Pb-212 (238.6 keV), Tl-208 (583.1 and 2614.7 keV) and Ac-228 (911.1 and 968.9 keV); four-week sealing to reach secular equilibrium; positive correlation between monazite grade and measured radioactivity. The paper does not state monazite's structural formula or describe thorium as structural rather than an impurity.
- UNVERIFIEDGAP — no nuclide-data or decay-scheme reference was available to this article, so nothing is asserted here about the structure of the thorium-232 chain, its step count or timings, the presence of radon-220 in it, or the field mechanisms (weathering, groundwater movement, surface leaching) by which secular equilibrium may be disturbed. An earlier draft asserted these; they are withdrawn pending a cited nuclide source.Non-public document · no public URL
- ANALYSISANALYSIS — the observation that a general mineral-specimen composition and a deposit's reported monazite percentage are different quantities on different bases, and that Osmond's published pairings of rare-earth oxide grade with monazite percentage for Orión imply a rare-earth content in the monazite above the general range quoted from Kotb et al., is this publication's framing rather than a claim made by Osmond or by any cited source. No conversion factor from assay to monazite percentage is published in Osmond's material, so no reconciliation is computed and no derived figure is stated. Osmond's mineral percentages are mass-balance estimates, not direct measurement; Orión has no JORC-compliant Mineral Resource or Reserve.Non-public document · no public URL
- ANALYSISANALYSIS — the framing that a radiometric anomaly is a proxy three inferential steps removed from a rare-earth grade, and that each step is a separate place the correlation can fail, is this publication's own way of organising the evidence above. It is not a claim made by either cited source.Non-public document · no public URL





