Heavy Metals in Chinese Herbs: Intentional Ingredient vs. Contaminant Limit

Heavy Metals in Chinese Herbs: Intentional Ingredient vs. Contaminant Limit

Two different things wear the same phrase

Search "heavy metals in Chinese herbs" and you get one worried blur. But underneath it are two completely different facts, and most sources slide between them without noticing:

1. Certain minerals were recorded in the classical materia medica as deliberate ingredients — mercury- and arsenic-bearing stones, added on purpose.
2. Modern pharmacopoeias set contaminant limits — ceilings on how much lead, cadmium, arsenic, or mercury may be present in a botanical product as unwanted impurity.

These are not the same claim, and conflating them produces nonsense in both directions. This article separates them, keeps each traced to its own source, and — this is the part worth staying for — shows that a single modern pharmacopoeia contains both regimes at once. Before anything else: this is a descriptive record of what classical texts and modern standards say. It makes no claim that anything is safe to consume, it recommends no intake of any kind, and it is not medical advice.

The classical record: minerals added on purpose

The classical texts did not treat mercury or arsenic as contamination to be screened out. They treated specific mineral forms of them as medicines, with their own entries, indications, and — startlingly to a modern eye — instructions to refine and ingest them.

Take cinnabar (dan sha). The Bencao Gangmu lists it among the top-grade entries of the Shennong Bencao Jing with the flavor line "sweet, slightly cold, non-toxic," and in its collected notes records the alchemists' esteem for it without irony:

仙方煉餌，最為長生之寶。

"In the immortals' recipes, refined and taken as a pill, it is the greatest treasure of long life." This is not a warning about a pollutant; it is a mineral prized enough to eat. (For why the same book then turns around and cites authorities calling cinnabar deadly — an argument inside the classical record — see our piece on cinnabar and mercury.)

Take realgar (xiong huang). The Shennong Bencao Jing, quoted in the Bencao Gangmu, gives it a full slate of indications and ends on the same alchemical note:

主寒熱，鼠瘻惡瘡，疽痔死肌，殺精物惡鬼邪氣百蟲毒，勝五兵。煉食之，輕身神仙。

"...it governs fevers and chills, scrofula and malign sores, ... it kills the hundred worm-poisons ... and refined and eaten, it lightens the body — one becomes an immortal." Again: an ingredient, by design, to the point of ingestion.

And the lead-bearing minerals — litharge (mi tuo seng), minium, and lead (qian) itself — sit in the mineral section of the same materia medica as recorded drugs, not as pollutants. Whatever we make of this today, the classical intent is unambiguous: these were ingredients, entered deliberately.

The modern chemistry: what those minerals are

Here is where the modern record speaks in a register the old texts could not. The composition of these prized minerals is now plain fact, not interpretation:

• Cinnabar is mercuric sulfide (HgS) — an inorganic mercury compound.
• Realgar is an arsenic sulfide (As₄S₄); its yellow cousin orpiment is As₂S₃.
• The lead minerals are lead compounds.

On what those elements do, the modern toxicology profiles are specific and public. The U.S. ATSDR Toxicological Profile for Mercury (NBK611017) records the nervous system and kidney as target organs of inorganic mercury. The ATSDR Toxicological Profile for Arsenic (NBK591633) records inorganic arsenic as a recognized human carcinogen, with documented effects including characteristic skin lesions, cancers, and cardiovascular effects. The profiles for lead (NBK589538) and cadmium (NBK158838) record the nervous system (and, for lead, neurodevelopment) and the kidney and bone, respectively.

Note what these paragraphs are and are not. They transcribe what modern authorities record about the elements and their compounds. They are not a verdict on what cinnabar or realgar do in any particular preparation, and they are not medical advice. The ancients had no concept of heavy metals; the classical entries could never have carried this meaning, and it sits on entirely different ground.

The modern standard: the same metals, as contaminants to limit

Now the second, separate fact. Modern pharmacopoeias and quality standards devote whole chapters to keeping these very elements out of botanical and pharmaceutical products — as contaminants, capped by numeric limits:

• The United States Pharmacopeia's General Chapters <232> (Elemental Impurities—Limits) and <233> (Procedures) govern elemental impurities in drug products, focusing on the four elements of greatest toxicological concern — arsenic, cadmium, lead, and mercury — and a separate chapter, <2232> (Elemental Contaminants in Dietary Supplements), sets limits for those same four elements in dietary supplements, botanicals included.
• The World Health Organization publishes guidance on contaminants in medicinal plant materials, recommending maximum permissible levels for heavy metals such as lead and cadmium.
• The Chinese Pharmacopoeia (2020) specifies a general standard for "heavy metals and harmful elements" (重金属及有害元素) — lead, cadmium, arsenic, mercury, and copper — with a dedicated general method (chapter 2321) for measuring them in botanical drugs.

In this frame the metals are the enemy: the point of the standard is to detect and cap them. This is the pathway most "heavy metals in herbs" alarm is really about — contamination, from soil, processing, or adulteration — and it is real. A widely cited body of work (Saper et al., JAMA, 2004 and 2008) found that a meaningful fraction of tested herbal products on the market contained detectable lead, mercury, or arsenic.

The irony: one pharmacopoeia, both regimes

Here is the detail that makes this a perfect case for source-traceability. The same Chinese Pharmacopoeia (2020) that caps mercury and arsenic as contaminants in botanicals also carries official drug monographs for cinnabar (朱砂, Cinnabaris) — defined as mercuric sulfide — and realgar (雄黄, Realgar) — defined as an arsenic sulfide. In one volume, arsenic and mercury are simultaneously the impurities to be limited and the declared principal components of listed medicines.

That is not a contradiction to be "resolved" by picking a side. It is two genuinely different regulatory objects: a contaminant limit on botanicals, and a monograph for a mineral drug. Keeping them apart — each with its own source — is the only way to state the situation without garbling it.

And crucially, the harm literature documents both pathways, which is exactly why they must be kept distinct. The contaminant pathway is the Saper JAMA findings above. The intentional-ingredient pathway is its own record: the 2008 JAMA study explicitly separated products where metals were added on purpose (the Ayurvedic rasa shastra tradition) from those where metals were contamination; and in the Chinese-medicine literature, cases of arsenic poisoning have been traced to the deliberate realgar ingredient itself (e.g. Forensic Science International, 2019). Two different causes, two different records — and flattening them into "heavy metals in herbs" hides which one is which.

How MateriaTrace handles it: ingredient and contaminant, never merged

So "heavy metals in Chinese herbs" arrives as several records that must not be collapsed:

1. the classical record — cinnabar and realgar as deliberate mineral drugs, even refined and eaten, each line traced to the Bencao Gangmu and the Shennong Bencao Jing;
2. the modern chemistry — HgS, As₄S₄, and the ATSDR target-organ profiles, each with its identifier;
3. the modern standard — contaminant limits for As, Cd, Pb, Hg (USP, WHO, ChP), which treat the metals as impurities to cap;
4. the irony in one book — a single pharmacopoeia holding both a contaminant limit and a mineral-drug monograph.

Here is how we hold that:

• The classical lines stay exactly as written — including "the greatest treasure of long life" and "refined and eaten, one becomes an immortal" — marked as the classical record, not modernized into a safety claim.
• The modern facts — composition, target organs, contaminant standards — sit in a separate layer, each carrying its own source, never edited into the classical text.
• We do not collapse them into a verdict. We do not tell you any mineral, at any level, is "safe" — that word appears nowhere in this article as our claim, and printing a permissible number would be exactly the intake guidance we refuse to give. We record what the classics intended, what the elements are, and what the standards specify — and keep them side by side.

Worth stating plainly: this article describes what classical texts recorded and what modern standards specify. It makes no claim that any mineral or any level of any metal is safe to consume, it gives no intake or dosage guidance, and it is not medical advice. Several substances here (cinnabar, realgar, the lead minerals) contain mercury, arsenic, or lead. Nothing in this article should be read as a recommendation to take them. MateriaTrace records what sources say and makes no treatment recommendation; any medical question belongs with a qualified professional.

Why the two records have to stay apart

It would be easy to print one scary line — "Chinese herbs contain heavy metals" — and stop. But that line fuses two facts that point in opposite directions: a tradition that added certain metal-bearing minerals on purpose, and a modern effort to keep metals out as contamination. You cannot understand a realgar poisoning case (intentional ingredient) with a contaminant-limit frame, or a lead-tainted botanical (contamination) with a monograph frame. The distinction only survives if each record keeps its own source — the Shennong Bencao Jing's "refined and eaten" next to ATSDR's target organs next to the pharmacopoeia's contaminant chapter — so you can see which fact you are actually looking at.

See the minerals themselves: cinnabar (dan sha) and realgar (xiong huang), recorded as mineral drugs; mercury (shui yin), what cinnabar yields when heated; and the lead minerals lead (qian) and litharge (mi tuo seng). For the argument the old texts had with themselves about cinnabar, read about cinnabar and mercury; for a toxic herb the classics processed on purpose, read about aconite and its processing.

If this "every record clicks back to its own source" approach is useful to you, see what else is in MateriaTrace on the homepage.

MateriaTrace is a source-traceable, bilingual knowledge graph of Chinese medicine. It records how classical and modern sources describe each entry — each field traceable to its primary source, contradictions kept side by side — and makes no treatment or usage recommendations. Every record in this article can be checked on its entry page.