M Potency · ICP-OES

ICP-OES for mineral ingredients

Technical note · Reviewed 18 November 2026 · All methods

ICP-OES measures the light emitted by atoms excited in an argon plasma, and it is used for elemental assays at percent levels and for multi-element panels. For a mineral powder it gives total magnesium and the other elements in the same digest. It is accurate and widely available, and it still cannot see the chemical form.

In short

A multi-element emission method that measures total magnesium at percent levels with a wide working range.

How the sample is prepared and measured

A weighed portion is digested in closed vessels with concentrated nitric acid, usually in a microwave system, until the powder is completely in solution. The digest is diluted to a fixed volume in dilute acid, and a set of matrix-matched calibration standards is run with the samples. Complete dissolution matters here: an undigested residue carries magnesium out of the measurement and shows up as a low result.

The solution is nebulised into an argon plasma at roughly 6,000–10,000 K, where the atoms are excited and emit at element-specific wavelengths. Magnesium is read at 279.553, 280.270 or 285.213 nm, with the line chosen to keep the signal inside the calibrated range. An internal standard such as yttrium or scandium corrects for drift and for differences in nebulisation efficiency.

A percent-level magnesium sample is diluted by a factor of about a thousand before it reaches the plasma, because the detector is linear over a limited range and the major line saturates easily. The dilution factor, the blank and the calibration curve are all part of the result. Panels for calcium, potassium, sodium, iron, zinc, copper and manganese can be run from the same digest.

What the result proves

ICP-OES gives a quantitative total for magnesium and for the other elements in the panel, and its working range covers both percent-level majors and low mg/kg minors. That range is why it is used as an assay method for mineral salts as well as a screen for other elements in the same material. Precision at percent levels is typically within a few percent relative.

When the ICP-OES assay and the EDTA titration agree within their stated uncertainty, the total magnesium figure is well supported. The result is also a suitable basis for a label claim, because it measures the element rather than a salt. It does not, by itself, say which compound supplied the element.

The same digest can confirm that the potassium, calcium or sodium content matches the specification, which matters when a chelate is blended with a carrier. That is a useful cross-check on the identity of the raw material, although it is not a substitute for a phase or spectroscopic test. It also shows whether a mineral level was adjusted with a second salt.

What it cannot tell you

The plasma destroys the sample, so the measurement is blind to the original species. Magnesium from a chelate, an oxide or a citrate salt produces the same emission line and the same percentage. ICP-OES cannot show coordination, the glycine ratio or the crystal phase.

Detection limits are the second limit. For lead, cadmium and mercury at the levels used for food supplements, ICP-OES is workable for lead but marginal for mercury at 0.10 mg/kg and unreliable for arsenic at low microgram-per-kilogram targets. A heavy-metal section run by ICP-OES rather than ICP-MS deserves a question about the limit of quantification.

Spectral overlap, high dissolved solids and matrix effects can bias individual lines, and a blocked nebuliser or a drifting plasma shows up as scatter rather than as an obvious failure. Refractory residues and incomplete digestion produce low results that look like a clean sample. The method also cannot detect contamination introduced after the digest is prepared.

What it proves

  • Total elemental magnesium at percent levels, with a working range that also covers minor elements.
  • A multi-element profile of the same digest, including calcium, potassium, sodium, iron, zinc and copper.
  • Agreement with an independent assay, when the two methods are run on the same lot.
  • The elemental figure used for a label claim or a specification comparison.

What it does not prove

  • The chemical form of the magnesium, or the ligand attached to it.
  • Trace contaminants at the low end of supplement limits, particularly mercury and arsenic.
  • Crystal phase, coordination or hydrate state.
  • Contamination introduced after digestion, or heterogeneity in the drum.

What to ask the laboratory

  1. Which emission line was used for magnesium, and was it checked for spectral overlap?
  2. Was the digestion closed-vessel, and what acid mixture was used?
  3. What is the limit of quantification for each element in this matrix?
  4. Is the instrument calibrated with matrix-matched standards and an internal standard?

Read the report for the emission wavelength, the digestion procedure and the dilution factor, because the percentage depends on all three. A magnesium result by ICP-OES should agree with the EDTA titration within the stated uncertainty, and a large gap points to incomplete digestion, a dilution error or a sample that was not representative. For lead, cadmium, mercury and arsenic at supplement limits, check that the laboratory used ICP-MS and reported a limit of quantification below the limit.

Where the method is used

The method pages sit behind the COA page. The terms used here are defined in the buyer glossary, and the regulatory limits are on the compliance page.

Send the certificate or specification you are reviewing and we will name the method behind each figure.

Request a specification review

Next step

A number without a method is not a result.

Send the certificate and we will name the method, the limit and the gap.