XRF Gemstone Analysis : What an X-Ray Beam Reveals About a Coloured Stone
Updated: Jul 18
The chromophores that give a gem its colour, the chemistry behind its value, and how to get a result you can trust !!
Somewhere in Bangkok, a rough reddish-purple tourmaline was placed in front of an X-ray fluorescence analyzer. In sixty seconds it returned a set of numbers. Whether those numbers meant anything depended entirely on one thing most people never think about: how the instrument had been told to interpret what it saw.
That single idea — that an XRF number is only as good as the assumption behind it — is one of the more useful hours a gemstone trader can spend. This article is that hour. We will start with what XRF actually does, work through what a gemstone looks like to an X-ray beam, read a real reading of this stone alongside an independent laboratory result, and finish with the handful of checks that separate a number you can trust from one that will cost you money.
XRF gemstone analysis allows traders to determine the chemical composition of gemstones within seconds without damaging the sample — but the results are only reliable if the instrument is properly calibrated for the material being measured.
What an XRF machine really sees
An X-ray fluorescence analyzer works on a simple and rather beautiful principle. It fires X-rays at your sample. The atoms inside absorb that energy, become briefly unstable, and then shed the excess by emitting X-rays of their own — at energies unique to each element. Copper always emits at 8.05 keV. Iron at 6.40. Manganese at 5.90. These are fingerprints, and they do not vary.
A detector counts those emitted X-rays and sorts them by energy, producing a spectrum: a chart with a peak wherever an element has announced itself. Copper's peak sits exactly where copper's peak must sit.
So far, so reliable. Identifying which elements are present is the easy half of the job — physics does most of the work.
The hard half is deciding how much of each. That requires the software to convert peak heights into percentages, and to do that it must make assumptions about what kind of material it is looking at. This conversion is called the calibration, and everything — everything — depends on it being right.
Here is the sentence to remember: Detecting elements is physics. Interpreting them requires context.
Why gemstone matrices require special consideration
Now consider what a tourmaline looks like to that beam.
Tourmaline is a borosilicate. Its structure is built mostly from boron, oxygen, sodium, lithium, aluminium and silicon. Those are light elements — low atomic number, few electrons, weakly fluorescing. To the conventional EDXRF field and benchtop screening instruments a trader actually uses, boron and oxygen are effectively invisible, and sodium and magnesium register only as faint whispers, detectable only under vacuum or a helium purge. (Specialised laboratory systems can reach further down the periodic table; the commercial screening analyzers in this story cannot.)
The elements a trader cares about — copper, manganese, iron — are the chromophores, the colouring agents. And in a real gemstone they are present at trace levels: fractions of a percent to a few percent, no more.
So an XRF looking at a tourmaline sees a stone whose overwhelming majority of mass is invisible to it, and a handful of faint heavy-element peaks representing a sliver of the whole.
What the software does next is the entire story.
The physics you can actually see. Look at a raw spectrum from one of these stones and you will notice, alongside the sharp element peaks, a large broad hump spread across roughly 18–33 keV. That is scattered tube radiation. Light elements don't absorb the beam efficiently — they scatter it. A solid metal produces almost no such hump. The scatter mound is the visible fingerprint of a light matrix: the instrument is not guessing that most of the stone is light material. You can read it directly off the spectrum.
The calibration mode that decides everything
Every calibration carries an assumption about the sample in front of it. For a gemstone, the most consequential one is how the analytical model treats the parts of the stone that contribute little or nothing to the measured spectrum.
A calibration built for solid metal — the kind used for karat and alloy testing — assumes that the elements it measures make up essentially the whole sample. For a gold ring, that assumption usually holds: gold, silver, copper, zinc and nickel are all readily detected, and together they are most of the material.
A tourmaline is a different problem entirely.
Most of its chemistry is light elements — boron, oxygen, lithium, sodium, aluminium and silicon. Depending on the detector, the measurement conditions and how the instrument is configured, some of these may be partially detected while others contribute primarily through matrix effects and scattering signatures rather than direct elemental peaks.. The elements a trader actually cares about — he chromophores, the colour-causing elements — are ones XRF can screen broadly across the transition-metal range: chromium (Cr), vanadium (V), titanium (Ti), manganese (Mn), iron (Fe), cobalt (Co), nickel (Ni), copper (Cu) and zinc (Zn) — the elements responsible for colour across a wide range of gem species, and typically present as only a small fraction of the whole stone.
Now the assumption turns against you. If the model takes the measurable heavy elements to represent the entire sample, then during normalization and matrix correction their reported concentrations can be pushed well above their true values — a small trace inflated into a large-looking percentage.
The physics underneath stays honest throughout. The characteristic X-ray peaks are real; they come from genuine elemental signals. What goes wrong is the translation — turning those signals into concentrations for a material whose chemistry is nothing like a metal alloy.
That is the heart of the matter. A gemstone read through a quantification model designed for metal can return numbers that simply do not reflect the chemistry of the stone — confident, precise, and unrepresentative all at once.
Getting it right depends on two things, and they solve two different problems.
The calibration determines the quantification. A model built for a mineral matrix — a dedicated gemstone calibration, a mineral or geological mode, or a fundamental-parameters model that accounts for light-element matrices and trace chromophores — does not force the measured elements to represent the whole stone. That is what keeps a trace copper reading as a trace rather than allowing it to become inflated during normalization and matrix correction.
The detector determines how much of the chemistry the instrument can directly observe. A large-area silicon drift detector, an ultra-thin window, and vacuum or helium measurement conditions all extend sensitivity toward lighter elements and improve precision and repeatability — allowing more of the stone's chemistry to be measured directly and less to be inferred from modelling assumptions.
Neither substitutes for the other. A high-performance detector operating under an inappropriate analytical model may still produce non-representative concentrations, while even the best calibration can only work with the information available in the measured spectrum.
Reliable gemstone analysis comes from both working together: a capable detector interpreted through a model that understands it is looking at a mineral rather than a metal alloy.
Because the objective was never simply to detect elements.
It is to interpret them within the chemical context of the stone itself.
Reading this stone correctly
So what is actually in the stone ?
Measured on the Elvatech ElvaX-JL, configured to account for the light-element matrix rather than force the total to one hundred percent, the reddish-purple tourmaline read like this:
Elvatech XRF Analyzer- Report
Element | Reported |
Copper (Cu) | 0.09% |
Manganese (Mn) | 0.05% |
Ga (Gallium) | 0.04% |
Iron (Fe) | 0.02% |
Cr (Chromium) | 0.02% |
Light elements (L.E.) | 99.78% |
Look at the last row, because it is the whole difference. The instrument states plainly that 99.78% of this stone is light material it cannot itemise — and reports the heavy chromophores as the traces they are. It does not pretend the invisible does not exist. It measures what it can, and declares the rest. That single honest line — the light-element balance — is what keeps the copper figure where it belongs: a fraction of one percent.
To confirm it, the same stone was sent to an independent, accredited gemological laboratory and analysed by LA-ICP-MS (laser ablation inductively-coupled-plasma mass spectrometry), a reference method that reports in parts per million.
The independent Gem Laboratory
An accredited gemological laboratory analysed the same stone by LA-ICP-MS (laser ablation inductively-coupled-plasma mass spectrometry), a reference method that reports in parts per million.
Element | Reported |
Copper (Cu) | 667.75 ppm = 0.067% |
Manganese (Mn) | 419.25 ppm = 0.042% |
Iron (Fe) | below detection limit |
The laboratory reports copper at 0.067%. The screening reading placed it at 0.09%. These are not identical figures, and they were not produced by equivalent methods — a screening XRF measurement on a rough, unpolished surface is not the analytical equal of laboratory LA-ICP-MS. But they describe the same practical reality: copper is present only as a trace constituent, well under one-tenth of one percent. The screening result is directionally consistent with the laboratory finding — which is exactly what a screening tool is supposed to deliver.
That is the quiet lesson of the whole article. A correctly configured reading did not just avoid an error. It told the truth about the stone, and an independent laboratory confirmed it.
Why the amount matters, not just the presence
A trader might reasonably shrug: the stone showed copper — I knew it was copper-bearing anyway.
That comfort is misplaced — and to see why, it helps to be clear about what actually earns a stone the name "Paraíba," because that is where the money is.
Here is the rule, stated plainly. The major laboratories — GIA, SSEF, Gübelin, GIT — do not classify Paraíba by a fixed copper number. There is no percentage a stone must cross to qualify. What they require is twofold: the right blue-to-green colour and saturation, and copper established as the cause of that colour. Copper must be present, and it must be doing the colouring work. A stone can contain copper and still not be called Paraíba if the colour is not there; and no figure on a screen confers the name if the laboratory does not see the colour. The name follows the colour — and copper must be its origin.
That is exactly why the amount of copper matters so much: not as a pass mark, but as the thing that produces the colour the name depends on. Copper concentration often correlates with saturation and with the characteristic neon appearance, though the final colour depends on several interacting factors rather than on copper alone — the concentrations of copper, manganese and iron, the oxidation states those ions adopt, any heat treatment the stone has undergone, and the stone's own optics. Copper is the lead actor, not the whole cast. Even so, where a stone's copper falls — a few hundred parts per million, or a few percent — is a first-order influence on whether it reads pale and ordinary or intense and extraordinary.
Manganese plays against it. Manganese contributes pink, red and violet tones, and the balance between copper and manganese helps explain the final hue — blue, green, or violet. That chemistry illuminates the colour mechanism of a stone; it is a window onto why the stone looks as it does. It does not, on its own, establish whether the stone has been heated. Treatment determination is a separate and more demanding question, addressed in a gemological laboratory through UV-Vis-NIR spectroscopy, FTIR, microscopic inclusion study, and the experienced interpretation of those combined results — not through the elemental chemistry a conventional XRF provides. The copper-to-manganese relationship is one clue gemologists may weigh when reasoning about heating, so a distorted set of figures muddies even that modest signal — but the verdict itself belongs to the laboratory, and no XRF reading, however clean, replaces it.
Now bring it back to the trading table, because this is where a wrong number becomes a wrong decision. If a reading tells you a stone is dense with copper when it is not, you have learned nothing you can use and a great deal you cannot. You cannot rank that stone against the next one. You cannot place it on the pale-to-neon scale that decides its value. You cannot reason about its colour mechanism or its treatment. And you cannot tell whether it is a stone that might one day reach for a premium name, or an ordinary one that never will. The single measurement that should inform the most important judgement you make — what this stone is worth — becomes fiction. And fiction, in a negotiation, always flatters.
"Copper is present" was never the valuable information. How much — and whether it is producing the colour — is what decides a stone's fate. In this case the representative chemistry tells a more modest story: copper is present only as a trace constituent, while manganese occurs at a comparable level. The chemistry is therefore more consistent with a reddish-purple copper-bearing tourmaline than with the intense copper-dominated blue-to-green colour mechanism typically associated with Paraíba-type material.
What an unreliable number costs, in practice
A number costs nothing to print, and a great deal afterwards.
A genuine stone gets rejected. It reads like an impossible metal alloy, the buyer grows uneasy, and a real opportunity walks out of the room. The stone was fine.
A stone is mispriced. Distorted copper and manganese figures destroy any reasoning about colour mechanism or relative quality. You overpay, or you undersell.
A supplier is wrongly accused. An impossible number can put honest material under suspicion — and, just as costly, the same unreliable reading can wave through the one stone that genuinely needed a second look.
Time bleeds away. Every questionable number becomes a re-test, a debate, a stone sent out to a laboratory to settle what good screening should have handled at the table.
Five Checkpoints for reliable XRF Gemstone Analysis
You do not need to become a spectroscopist. You need six habits.
Know what mode the machine is in. Analytical modes intended for precious metals and engineering alloys are optimized for metallic matrices. Coloured gemstones generally require mineral-oriented calibrations or quantification models that account for light-element matrices and trace chromophore elements.
Look for the light-element figure. For a silicate gemstone it should be high — commonly 90% or more. If a stone reads as nearly 100% heavy metals with no light-element balance, the software has force-normalized, and every percentage on that screen is inflated.
Ask whether the chemistry is possible. Results should always be assessed in the context of mineral chemistry and geological plausibility.
Demand repeatability. Measure the same spot three times. Representative measurements repeat within a narrow margin.. Numbers that wander are telling you something.
Use XRF as a screening tool, not as the final word. A properly configured measurement can provide valuable chemical information in seconds and help guide decisions about further investigation. Species identification, geographic origin, treatment determination and trade nomenclature remain the domain of specialist gemological laboratories using reference analytical methods.
In practice, XRF and laboratory analysis are complementary tools rather than competing ones. A gemstone trader often benefits from both: rapid, non-destructive screening at the point of decision, followed by laboratory confirmation when higher levels of certainty are required.
So — Should a gemstone trader use XRF ? — Yes.
Everything above is an argument for reading numbers wisely, not against XRF. Used correctly, X-ray fluorescence is one of the most useful tools a gemstone trader can own: it is fast, it is non-destructive, and in seconds it can tell you whether a stone's chemistry fits its story — whether the copper that should be there is there, whether something that shouldn't be present is raising a flag. That is real, everyday value. The failure this article describes is not a failure of XRF as a technique; it is what happens when a gemstone is measured through a model built for metal instead of one built for a mineral.
Two things make that everyday value dependable. The first is the calibration: a mineral-appropriate analytical model that reports a light-element balance and does not force a stone's traces to add up to the whole. The second is the detector: a capable modern detector — a large-area silicon drift detector with a thin window, and vacuum or helium conditions — can measure some of the lighter elements directly and improve precision, so more of the stone's chemistry is observed rather than inferred. Together they are what turn a fast reading into a trustworthy one.
Good gemstone screening then comes down to three habits this article has already given you: use a setup built for the material, not for metal; take more than one measurement and trust the numbers that repeat; and treat the result as a screening gate, not a laboratory certificate. Do that, and XRF earns its place on the desk.
The right tool for the job
Which analyzer suits you depends on how and where you work. For coloured-stone and jewellery screening, ITechSphere supplies the Elvatech range — including three natural starting points:
ElvaX Lab — Benchtop XRF
A benchtop laboratory analyzer built for controlled, repeatable measurement, designed for gemstone laboratories, jewellery manufacturers, and traders who need bench-grade certainty rather than field speed. Its enclosed measurement chamber, stable geometry, and camera-assisted positioning let you place the beam precisely on a small stone or a specific zone, while vacuum or helium measurement conditions extend its reach toward the lighter elements — the setup that keeps a trace chromophore reading honest. It is particularly suited to careful trace-element work on the chromophores that decide a coloured stone's identity and price — chromium (Cr), vanadium (V), iron (Fe), titanium (Ti), manganese (Mn), copper (Cu), cobalt (Co) and nickel (Ni) — across faceted and rough coloured stones, and any analysis where a valuable sample must not be altered. The measurements in this article were performed on an ElvaX Lab configured for gemstone screening and chromophore analysis.
ProSpector 3 Advance — Handheld XRF
A handheld analyzer designed for professionals who need to bring the instrument to the material rather than the material to the instrument. Built around a large-area silicon drift detector for strong sensitivity and fast throughput, it delivers laboratory-style analytical depth in a field-portable form. Its broader element range and refined light-element performance make it well suited to buying trips, parcel screening, exhibitions, and mixed gemstone-and-precious-metal workflows — screening the chromophores that drive value, such as chromium (Cr), iron (Fe), titanium (Ti), vanadium (V), manganese (Mn), copper (Cu) and cobalt (Co), and verifying karat and alloy composition on finished pieces in the same session.
ProSpector 2 LE — Handheld XRF
A portable, lightweight solution for rapid, non-destructive screening where mobility and speed matter most. It is built for a fast, reliable first-pass read: sorting parcels, checking incoming goods, and making go/no-go decisions at the table before a stone is worth a closer look. In seconds it can flag the key colouring elements — iron (Fe), chromium (Cr), manganese (Mn), copper (Cu) and more — across everyday gemstone and precious-metal evaluation where a benchtop unit isn't practical.
Whichever platform you choose, the principle remains the same: the analytical approach must match the material being measured.
For coloured gemstones, that means combining appropriate detector capability, measurement conditions and calibration models designed for mineral matrices and trace chromophore elements.
When these work together, XRF becomes what it is intended to be — a fast, non-destructive and highly valuable source of chemical insight for the gemstone trade.
The bottom line
Detecting elements is physics. Interpreting them requires context.
The tourmaline examined in this article contained trace copper and manganese levels that were directionally consistent with independent laboratory analysis, confirming the presence of copper as a minor constituent of the stone.
The lesson is not about one instrument or one method.
It is about understanding the sample, selecting the appropriate analytical setup, and interpreting the results within the context of the material being tested.
XRF remains one of the most valuable tools available for rapid, non-destructive gemstone screening.
Used with the right methodology, it provides meaningful insights in seconds and helps guide decisions that can later be supported by laboratory analysis when required.
Because gemstones are not alloys.
And every material tells its story differently.
Looking for a handheld or benchtop XRF analyzer for gemstone screening, precious-metal and karat testing !
ITechSphere supplies and supports the full Elvatech XRF range in Thailand — including the benchtop ElvaX Lab for gemstone and jewellery laboratories and the ProSpector handheld series for field and industrial work — backed by local calibration, training, and service. Whatever you test, we help you set the instrument up correctly for your material, so the numbers you rely on are numbers you can trust.
ITechSphere Co., Ltd. — authorized Elvatech distributor in Thailand.
.webp)
Comments