ICP-MS vs. ICP-OES: How to Choose the Right Heavy Metal Testing Method for Supplements and Cosmetics

Most brands ordering heavy metal testing don’t realize they’re choosing between two fundamentally different analytical technologies. And the wrong choice doesn’t just waste money — it can produce compliance data that doesn’t actually answer the regulatory question you’re trying to settle.

Both ICP-MS (Inductively Coupled Plasma – Mass Spectrometry) and ICP-OES (Inductively Coupled Plasma – Optical Emission Spectrometry) use a high-temperature argon plasma to atomize and ionize your sample. After that, the instruments diverge completely. ICP-MS separates ions by mass-to-charge ratio; ICP-OES measures the wavelength of light emitted when excited atoms return to their ground state. That single engineering decision cascades into dramatically different performance profiles — and different regulatory fit.

Here’s how to think through the decision.

Detection Limits: Where the Real Difference Lives

ICP-MS routinely achieves method detection limits in the parts-per-trillion (ppt) range — as low as 0.001 µg/L for elements like lead or arsenic in a prepared digest. ICP-OES sits roughly three orders of magnitude above that, with practical detection limits typically in the 1–50 µg/L range depending on the element and matrix.

For most regulatory thresholds that matter right now, that gap is decisive. USP Chapters 232 and 233 — the framework governing elemental impurities in pharmaceutical-grade products — set a permitted daily exposure (PDE) for lead in oral products at 5 µg/day. For arsenic, the oral PDE is 15 µg/day. California Proposition 65’s maximum allowable dose level (MADL) for lead goes further: just 0.5 µg/day. Translated into a concentration limit for a typical two-capsule supplement serving, that works out to low single-digit ppb values.

At those thresholds, ICP-OES may not deliver the sensitivity needed to confidently report “below limit.” You’ll get a number, but the measurement uncertainty around it can be wide enough to leave the compliance determination genuinely ambiguous. ICP-MS closes that gap.

That said, plenty of testing scenarios don’t require ppt sensitivity. If you’re qualifying a bulk botanical raw material against a specification of 10 ppm lead, and the matrix is relatively clean, ICP-OES is entirely appropriate — and typically faster, with less sample preparation overhead.

Matrix Complexity and Spectral Interference

Here’s something that doesn’t appear in instrument spec sheets: both technologies have matrix-dependent failure modes, and understanding them is half the work of method selection.

ICP-MS can be plagued by polyatomic interferences — situations where two ions combine to produce a mass-to-charge ratio that mimics the element you’re measuring. The textbook example: ⁴⁰Ar + ³⁵Cl produces an ion at 75 m/z, which overlaps directly with arsenic-75 (⁷⁵As). In a high-chloride matrix — marine-derived supplements, electrolyte products, or samples digested with hydrochloric acid — this interference can artificially inflate arsenic readings unless the instrument uses a collision/reaction cell (CRC) with helium or hydrogen mode, or the method applies a validated correction algorithm.

Modern ICP-MS instruments with CRC technology largely solve this problem. But it’s worth asking your lab specifically how they handle arsenic in chloride-rich matrices. An ISO 17025-accredited lab should have documented validation data demonstrating interference correction performance for the matrix types they accept.

ICP-OES has its own interference landscape: spectral overlap from emission lines of co-eluting elements, and matrix-induced plasma temperature shifts that move signals. Cosmetic matrices with high calcium, magnesium, or titanium content — common in mineral sunscreens and foundation formulations — can suppress or enhance signals for other target analytes. Experienced analysts address this with matched internal standards and background correction algorithms, but it requires thoughtful method development, not just loading the sample and running the default parameters.

The practical conclusion: neither technology is universally superior without knowing what’s actually in your product matrix.

Regulatory Frameworks That Drive the Method Choice

Different product categories carry different regulatory sensitivity expectations, and those expectations were written with specific detection capabilities in mind.

Dietary Supplements (21 CFR Part 111, USP 232/233): FDA’s cGMP regulations don’t mandate a specific analytical method, but USP Chapter 233 specifies that validated methods must achieve a limit of quantitation (LOQ) at or below 30% of the applicable specification limit. For lead and arsenic in products subject to Prop 65 exposure, that performance requirement often means ICP-OES simply can’t clear the bar. ICP-MS is the only instrument with enough sensitivity to satisfy it.

Cosmetics (21 CFR Part 700, MoCRA): FDA’s 2016 guidance on lead in lip products suggested action levels in the low single-digit ppm range. The Modernization of Cosmetics Regulation Act (MoCRA), signed in December 2022, has added product safety substantiation requirements being phased in through FDA rulemaking. For lip products and products applied near the eyes, where lead action levels sit at or below 10 ppm and Prop 65 exposure calculations push the compliance threshold into the ppb range, ICP-MS is the appropriate choice.

Foods (FDA “Closer to Zero”): FDA’s Closer to Zero action plan, targeting lead, arsenic, cadmium, and mercury in foods marketed to infants and young children, has proposed action levels as low as 10 ppb for lead in certain baby food categories. At 10 ppb — with a target LOQ of 3 ppb under the USP-equivalent 30% rule — ICP-MS isn’t just preferable. It’s functionally required to generate defensible compliance data.

Prop 65 (California OEHHA): With a lead MADL of 0.5 µg/day and an inorganic arsenic MADL of 10 µg/day, Prop 65 compliance thresholds for most product categories translate to per-serving concentration limits in the 1–20 ppb range. There’s no version of that calculation where ICP-OES is the right tool.

When ICP-OES Is Actually the Better Choice

It’s tempting to read the above and conclude “just always use ICP-MS.” That’s not the right answer, and any lab that doesn’t push back on that instinct a little isn’t giving you the full picture.

ICP-OES handles high-concentration matrices and macro-element quantification more cleanly than ICP-MS. If your product’s specification requires measuring calcium at 200 ppm, zinc at 50 ppm, and iron at 30 ppm — as is common in mineral supplements and fortified foods — ICP-OES is faster, offers a wider linear dynamic range in the high-ppm regime, and typically delivers meaningfully lower cost per sample. ICP-MS detectors saturate at high analyte concentrations, requiring significant dilution that introduces its own sources of uncertainty.

For raw material incoming qualification — screening a dozen elements across a botanical ingredient panel where the thresholds are in the tens-of-ppm range — ICP-OES gets you actionable data without the cost premium. If you’re running 200 incoming lots per month, that difference compounds quickly.

And for certain regulated products where FDA guidance or USP monograph limits sit at 1 ppm or above, ICP-OES’s sensitivity is entirely sufficient. The key is always matching the method’s LOQ to the regulatory threshold it needs to address.

The Case for Running Both

Here’s the option that surprises some clients: you don’t always have to choose. For products that require both macro-element quantification (zinc, calcium, magnesium, copper at specification levels) and trace-element compliance verification (lead, arsenic, cadmium, mercury at ppb levels), a well-designed testing package can deploy ICP-OES for the high-concentration panel and ICP-MS for the trace-element compliance panel on the same digest aliquot.

That approach is typically more cost-efficient than forcing a full ICP-MS panel to handle high-concentration analytes with dilution workarounds, or asking ICP-OES to reach detection limits it wasn’t designed for. We use this combination routinely for mineral supplement brands and multi-ingredient personal care products — the extra coordination during method setup pays for itself in cleaner data.

Five Questions Worth Asking Your Testing Lab

Before you submit samples, these questions will tell you quickly whether a lab’s heavy metal testing program is technically sound:

1. What is your reported LOQ for lead and arsenic in this matrix, and how does it compare to the applicable regulatory threshold? If the gap is less than 3× the LOQ, the method is operating in an uncomfortable zone.

2. Is this method validated for my specific product matrix? A dietary supplement capsule, a lipstick, and a baby food puree each behave differently during acid digestion. Ask to see validation summary data for a comparable matrix.

3. Is the lab ISO 17025 accredited for heavy metals in this product category? Accreditation isn’t a guarantee of accuracy, but it’s the baseline indicator that an external assessor has reviewed the lab’s procedures, equipment calibration records, and analyst competency.

4. How do you handle polyatomic interferences for arsenic in chloride-containing samples? The answer should reference CRC gas mode, reaction cell chemistry, or a validated mathematical correction. Vague answers here warrant follow-up.

5. Which internal standards does your method use, and how are they matched to target analytes? A rigorous ICP-MS method uses multiple internal standards selected by mass and ionization potential to bracket the target analytes. This is the quality control that catches matrix suppression before it affects your result.

The lab that answers all five without hesitation — and can back up those answers with method validation documentation — is almost certainly producing more defensible data than one that treats these questions as unusual.

Getting a number back isn’t the same as getting compliance-ready data. For products where a Prop 65 notice or an FDA inquiry is a real possibility, the difference between those two things is exactly what the method choice determines.

---

Written by Nour Abochama, Vice President of Operations, Qalitex Laboratories. Learn more about our team

Talk to our team about your testing needs. Contact us

Related from our network

• Understanding FDA Regulatory Requirements for Analytical Methods — Aurora TIC covers FDA audit readiness and 21 CFR compliance strategy for regulated industries.
• Heavy Metal Testing for Raw Material Supplier Qualification — Ayah Labs specializes in COA verification and incoming raw material testing for B2B manufacturers.