Heavy-metal testing: ICP-MS vs. ICP-OES for botanical extracts

Heavy-metal contamination is one of the few analytical lines on a botanical extract Certificate of Analysis where the regulatory limits across the major export markets are tight enough to drive method choice. ICP-MS and ICP-OES — the two working techniques — are not interchangeable at the regulated specification levels, and the choice of method has consequences that propagate through to the downstream framework. This note sets out the working comparison.

What ICP-MS and ICP-OES actually do

Both ICP-MS (inductively-coupled plasma mass spectrometry) and ICP-OES (inductively-coupled plasma optical emission spectroscopy) use a high-temperature argon plasma to atomise and excite the elements in a sample. The methods diverge after the plasma. ICP-OES measures the wavelength-specific light emitted by the excited atoms; ICP-MS sorts the ionised atoms by mass-to-charge ratio and counts them at the detector. The fundamental difference is what the detector measures and at what sensitivity.

ICP-OES is robust, well-understood, and entirely adequate for elemental quantification at concentrations down to roughly the parts-per-million level. ICP-MS reaches three to four orders of magnitude lower — parts per billion in the working range, parts per trillion at the limit. The methods have different cost profiles, different matrix tolerances, and different operational workflows.

Where ICP-OES is the right answer

For botanical extracts whose specifications sit at the parts-per-million level — and many do — ICP-OES is the working choice. The technique is widely available, the per-batch cost is lower, and the analytical robustness is well-understood. For lead, cadmium, arsenic, and mercury at specifications around 1–10 ppm or higher, ICP-OES is straightforwardly fit for purpose.

ICP-OES is also more tolerant of high-matrix samples. Botanical extracts can carry significant organic matrix, and the optical detection of ICP-OES is less affected by matrix interferences than mass-spectrometric detection. For some matrix categories — high-mineral-content roots, certain fermented products — ICP-OES is the more reliable workflow even where ICP-MS sensitivity is not strictly required.

Where ICP-MS is the right answer

ICP-MS becomes the working choice when the regulatory limit on the relevant element is below the parts-per-million range, or when the buyer's framework requires quantification of multiple trace elements across orders of magnitude in a single run. The EU's maximum limits for arsenic and lead in some food categories sit in the low-ppb range; the FDA's interim reference levels for cadmium and lead in food are similarly tight. At these levels ICP-OES cannot meaningfully discriminate signal from baseline.

ICP-MS is also the working choice for elemental impurity testing under ICH Q3D — the international guideline for elemental impurities in pharmaceutical products. Q3D establishes permitted daily exposure limits for elements at concentrations that ICP-OES does not reach. Suppliers producing botanical extracts intended for pharmaceutical or pharmaceutical-adjacent use need ICP-MS in the analytical chain or through the partner laboratory matrix.

Speciation

The harder analytical question on heavy metals in botanicals is speciation — particularly inorganic versus organic arsenic. Total arsenic in many botanicals will exceed the inorganic-arsenic regulatory limit by an order of magnitude or more, because organic arsenic species predominate. Inorganic arsenic is the toxicologically relevant species; total arsenic is the easy measurement.

Speciation analysis — typically hyphenated HPLC-ICP-MS — quantifies inorganic and organic arsenic separately. Where the regulatory limit is on inorganic arsenic and the total arsenic reading is at or above the limit, speciation is the working response. The cost is higher than total-element analysis; the result is the one the regulator actually requires.

Jurisdictional differences

The major export markets approach heavy-metal limits differently. The EU sets explicit maximum levels for some heavy metals in some food categories under Regulation (EC) 1881/2006 as amended, with specific limits on inorganic arsenic in some contexts. The FDA uses interim reference levels and category-specific guidance rather than blanket maxima. China's GB standards apply by product category. The ASEAN region operates a harmonisation framework that varies in implementation.

For an ingredient supplier shipping into multiple regulated markets, the working specification on the COA needs to satisfy the tightest applicable limit across the buyer's intended export markets — and the method needs to support that limit with adequate margin. ICP-MS sets the working method for ingredients intended for multi-market premium-tier supply; ICP-OES remains adequate for regional supply at less restrictive specifications.

What this means for the COA line

A heavy-metal line on a Certificate of Analysis should specify the method used, the detection limit of the method, the elements measured, and — where relevant — the speciation result. A COA that lists "heavy metals: within limits" without method, detection limit, or per-element result is not adequate for any regulated-market framework and is the signal of a supplier whose analytical infrastructure is incomplete. The working COA line names ICP-MS or ICP-OES, lists arsenic (with speciation where applicable), cadmium, lead, and mercury individually, and quantifies each against the relevant regulatory limit.