Why ICP-MS Is the Only Method That Satisfies USP <232>/<233> for Herbal Supplement Heavy Metals

Every batch of ashwagandha, turmeric, or sea kelp arriving at your receiving dock is a bioaccumulator. That’s not alarmism — it’s plant physiology. Botanical root systems and cell walls concentrate whatever elemental profile exists in the surrounding soil, and for ingredients sourced from South Asia, East Asia, or South America, that profile regularly includes lead, cadmium, arsenic, and mercury at concentrations that matter to both regulators and the consumers reading your label.

A 2013 study published in JAMA Internal Medicine tested 193 Ayurvedic herbal products sold online and found that 20.7% contained detectable lead, mercury, or arsenic. More recent ConsumerLab surveillance data has flagged heavy metal exceedances in turmeric extracts, chlorella powders, and several popular adaptogen products. The common thread across most of those cases: the supplier CoA showed a “pass” result — because the supplier used a test method that couldn’t see what was actually there.

Getting heavy metals right starts well before your production release. It starts with understanding what USP <232> and <233> actually require, and why ICP-MS isn’t just the preferred analytical method — it’s the only one that fully satisfies the procedure as written.

What USP <232> Actually Specifies: PDEs, Not Generic ppm Limits

Most quality teams know USP <232> as “the heavy metals chapter,” but the chapter is more precise than that framing suggests. It doesn’t set blanket parts-per-million cutoffs. It establishes permitted daily exposure (PDE) limits tied to route of administration, which means the acceptable concentration in your raw ingredient depends directly on how much of that ingredient a consumer takes each day.

For the four Class 1 elements — the highest toxicological concern — the oral PDEs are:

• Lead (Pb): 5 µg/day
• Arsenic (As): 15 µg/day
• Cadmium (Cd): 5 µg/day
• Mercury (Hg): 30 µg/day

To convert these into a usable concentration limit for your raw material specification, you divide each PDE by the maximum daily intake of that ingredient in your formula. A product where consumers take 3,000 mg of turmeric powder per day has a very different lead specification limit than one where turmeric appears at 50 mg per softgel. That calculation — not a supplier template’s generic “NMT 10 ppm” entry — is what your specification should reflect.

There’s a second layer that trips up a lot of Midwest brands distributing nationally: California’s Proposition 65. The Maximum Allowable Dose Level (MADL) for lead under Prop 65 is 0.5 µg/day — ten times tighter than the USP <232> PDE. For inorganic arsenic, it’s 10 µg/day. For any brand selling through e-commerce or into retail with California exposure (which is essentially every national brand), Prop 65 is the effective floor, not USP <232>. Your analytical testing lab needs to report at a sensitivity level that lets you evaluate compliance against both standards simultaneously.

Why ICP-MS Is the Reference Procedure for USP <233>

USP <233> describes the actual test procedures for elemental impurity measurement. It accepts two instrumental techniques: ICP-OES (inductively coupled plasma optical emission spectrometry) and ICP-MS (inductively coupled plasma mass spectrometry). Both are listed. But the reality of herbal ingredient matrices — and the tight specification limits that fall out of Prop 65 MADL calculations — pushes virtually every compliant contract lab toward ICP-MS.

Here’s why the detection capability gap is decisive.

ICP-OES typically achieves instrument detection limits in the range of 1 to 50 µg/L in solution, depending on the element and matrix. ICP-MS routinely reaches 0.001 to 0.1 µg/L — roughly 100 to 1,000 times lower. When you’re evaluating a botanical ingredient against a specification limit derived from a 0.5 µg/day Prop 65 MADL at a 1–3 gram daily dose, the resulting concentration limit for lead lands somewhere between 0.17 and 0.5 ppm. At that level, ICP-MS isn’t a preference. ICP-OES can’t generate a reliable quantitative result at the concentration that actually matters.

There’s also the matrix interference problem. Dark botanical powders — ashwagandha root, valerian, elderberry, black cohosh — produce complex acid digestion matrices that create spectral overlaps in ICP-OES. ICP-MS separates ions by mass-to-charge ratio rather than optical emission, which handles many of these interferences inherently. Modern collision/reaction cell (CRC) technology has further addressed polyatomic interferences on critical elements: the ArCl⁺ interference on ⁷⁵As, for example, is a well-documented challenge in herbal matrices that CRC-equipped instruments resolve reliably.

What about older methods? AAS (atomic absorption spectrometry), colorimetric testing, and portable XRF screening are not compliant USP <233> procedures for finished CoA reporting. Colorimetric heavy metals tests can’t quantify individual elements and don’t meet the precision requirements of the current chapter. AAS requires separate instrument runs per analyte and typically can’t match ICP-MS detection limits for the low-concentration compliance work that Class 1 elements now demand. Portable XRF has legitimate utility for incoming screening decisions, but any CoA based solely on XRF data should prompt an immediate supplier conversation about their testing program.

Three CoA Red Flags That Signal an Inadequate Heavy Metals Test

After reviewing CoAs from suppliers serving dozens of Midwest supplement brands, we see inadequate heavy metals testing show up in predictable patterns. These three flags should pause your receiving process:

1. Results reported as “ND” with no stated detection limit. “Not detected” without a quantified reporting limit is analytically meaningless. If a supplier’s instrument has a detection limit of 2 ppm for lead but your specification requires ≤0.5 ppm, a “not detected” result tells you nothing about compliance. A properly formatted CoA states the method detection limit (MDL) or reporting limit for each element — and that limit must fall below your specification limit, not above it.

2. A single “total heavy metals” result instead of element-specific values. USP <232> requires individual PDE evaluations for each of the four Class 1 elements. A CoA reporting “total heavy metals: pass at NMT 10 ppm” is referencing a legacy format from USP general chapter <231>, which was officially retired in 2018. If your supplier is still issuing results in that format, their testing program hasn’t been updated in at least seven years — and that raises legitimate questions about everything else in their quality system.

3. No method reference or digestion parameters. ICP-MS analysis of botanical matrices requires microwave-assisted acid digestion at high pressure and temperature. The dilution factors involved — typically 100:1 or higher in the final solution — must be documented and traceable for independent verification. At minimum, a compliant result report should reference the procedure used (e.g., USP <233>, Procedure 1), the instrument model, and the calibration traceability. Results without analytical parameters can’t be verified and shouldn’t support a release decision.

What Compliant ICP-MS Reporting Looks Like in Practice

When Midwest clients ship botanical raw materials to our Countryside, IL receiving facility outside Chicago, samples route to our ISO 17025–accredited analytical testing laboratory in California within 24 to 48 hours. USP <233> ICP-MS panels cover all four Class 1 elements as standard, with Class 2A elements (cobalt, nickel, vanadium) available for formulas where those contaminants are relevant to the ingredient profile.

For high-risk ingredients — spirulina, chlorella, rice protein, kelp — we offer arsenic speciation to separate inorganic arsenic (the regulated form under Prop 65) from organic arsenobetaine, which is predominantly present in marine-sourced ingredients and carries a much lower toxicological concern. That distinction matters: a total arsenic result of 1 ppm in a spirulina powder looks alarming against a 0.17 ppm Prop 65-derived limit, but if 95% of that arsenic is organic arsenobetaine, the inorganic arsenic concentration may be well within compliance. Speciation is the only way to make that determination.

A compliant CoA from an accredited lab includes: element-specific concentrations in µg/g or µg/kg, method detection limits for each element, the digestion method reference, instrument model and calibration traceability, and the ISO 17025 accreditation scope. When results arrive in that format, you have everything needed to calculate compliance against USP <232> PDEs at your finished-product daily dose — and against Prop 65 MADLs if California is part of your distribution footprint.

Standard turnaround from Chicago sample receipt to final CoA is 5 to 7 business days. Rush options are available for time-sensitive production releases.

Build the Specification First, Then Test

Getting the right analytical method is half the equation. The other half is knowing what limit you’re actually testing against before the sample ever ships to a lab.

If your raw material specification lists “lead: NMT 10 ppm” because that’s what the supplier template defaulted to, but your actual daily dose calculation yields a required limit of 1.2 ppm, you’ve run a test that doesn’t demonstrate compliance — even if the result comes back clean. The math isn’t complicated: take each Class 1 PDE, divide by the maximum daily dose of that ingredient in your formula, and you have your concentration limit. Do it for all four elements. Compare each against the corresponding Prop 65 MADL if you have national distribution. The lower of the two values is your specification limit.

That calculation takes about 30 minutes and is the foundation of a quality program that would survive an FDA inspection or a third-party audit. It also tells you exactly what detection capability to require from your analytical testing lab before you approve them as a supplier. Most quality failures in incoming raw material testing aren’t failures of the instrument — they’re failures of the specification. Fix the specification, and the rest of the program follows.

---

Written by Nour Abochama, VP Operations, Qalitex | Quality Consultant, Ayah Labs. Learn more about our team

Ship your sample to our Chicago facility — get a Qalitex CoA in 5–7 days. Contact us

Related from our network

• Supplement Raw Material Testing & ISO 17025 Accreditation — Full-scope analytical testing laboratory for dietary supplement ingredients, botanicals, and finished products under ISO 17025 accreditation.