Total arsenic determination is a fundamental safety test for Chinese herbal medicines, ensuring compliance with pharmacopoeial heavy metal limits and protecting consumers from carcinogenic elemental contaminants. This technical guide explains the principal analytical methods—including hydride generation atomic fluorescence spectrometry (HG-AFS), atomic absorption spectrometry (AAS), and inductively coupled plasma mass spectrometry (ICP-MS)—together with sample digestion procedures, method validation requirements, and regulatory limits. It is intended for laboratory analysts, quality assurance professionals, and herbal drug manufacturers involved in routine testing and regulatory submission.
Technical Overview
Arsenic is a ubiquitous metalloid that enters Chinese herbal medicines primarily through the geochemical background of cultivation soils, contaminated irrigation water, agrochemical inputs, industrial emissions, and, in rare historical cases, the deliberate use of arsenic-bearing minerals during processing. Because inorganic arsenic species, namely arsenite As(III) and arsenate As(V), are classified as Group 1 human carcinogens by the International Agency for Research on Cancer (IARC), even low-level chronic exposure poses serious health risks, including skin lesions, peripheral neuropathy, hepatotoxicity, and cancers of the skin, bladder, and lung.
Regulatory authorities worldwide have therefore established strict limits for total arsenic in herbal drugs. The Chinese Pharmacopoeia (2020 edition) specifies that total arsenic in most herbal materials shall not exceed 2 mg/kg unless otherwise stated in individual monographs, and comparable requirements appear in the United States Pharmacopeia elemental impurities standards, the Japanese Pharmacopoeia, the Hong Kong Chinese Materia Medica Standards, the WHO guidelines on herbal material quality, and the Green Trade Standards for Medicinal Plants and Preparations. In addition, a small number of traditional formulations legitimately contain realgar (arsenic sulfide), which makes rigorous arsenic monitoring essential for both routine safety screening and product-specific risk assessment.
A complete determination workflow comprises representative sampling, milling and homogenization, decomposition of the organic matrix through acid digestion, instrumental quantification, and validation-supported reporting of results on a dry weight basis. Each stage introduces potential sources of error, and a robust method must control contamination, analyte loss, and spectral interference throughout.
Testing Samples
Crude Herbal Materials (Botanical Drugs)
Whole or cut dried botanical drugs, including roots, rhizomes, barks, leaves, flowers, fruits, seeds, and whole herbs, represent the majority of submitted samples. Because arsenic distribution in plant tissue is heterogeneous, a representative sample of 50 to 100 g is typically collected by quartering, dried to constant weight, ground to pass a 40- to 60-mesh sieve (approximately 0.25 to 0.45 mm), and stored in acid-washed polyethylene containers to prevent cross-contamination.
Processed Herbal Slices and Decoction Pieces
Slices produced by washing, cutting, steaming, stir-frying, honey-frying, or carbonizing may carry arsenic introduced by both the raw material and the processing aids, including water, excipients such as honey or vinegar, and contact surfaces. Testing of processed slices is required because pharmacopoeial limits apply to the material as consumed in clinical decoctions.
Herbal Extracts, Concentrates, and Granules
Concentrated water or hydroalcoholic extracts and spray-dried granules are tested both for pharmacopoeial compliance and for process control. Extraction can either concentrate arsenic relative to the crude drug or reduce it if the element partitions poorly into the extracting solvent, so a mass-balance comparison between raw material and extract is often performed during method development.
Finished Herbal Preparations
Pills, tablets, powders, capsules, and oral liquids containing single herbs or multi-herb formulas are tested as finished products. For preparations containing known arsenical ingredients such as realgar, total arsenic determination is combined with speciation analysis to quantify the toxic inorganic fraction and to support dose-based risk assessment.
Testing Items
The core deliverable of the analysis is the total arsenic content, expressed in mg/kg (equivalent to ppm) on a dry weight basis. A complete test report typically covers the following items:
- Total arsenic concentration in the as-received sample and converted to dry weight using the measured moisture or loss-on-drying value
- Method blank, calibration verification, duplicate digestion results, and spike recovery for the analytical batch
- Detection limit and quantitation limit achieved for the specific matrix and dilution factor
- Comparison of the result against the applicable pharmacopoeial or contractual limit, with a clear pass or fail conclusion
- Where required, arsenic speciation (As(III), As(V), monomethylarsonic acid, dimethylarsinic acid, arsenobetaine) by HPLC coupled to ICP-MS, particularly for marine-derived or algal herbal materials in which total arsenic may overestimate toxicological risk
Testing Methods
Sample Digestion and Decomposition
Microwave-assisted acid digestion is the preferred decomposition technique. A test portion of 0.2 to 0.5 g, weighed to the nearest 0.1 mg, is treated with 5 to 8 mL of high-purity concentrated nitric acid, often with 1 to 2 mL of hydrogen peroxide added to complete the oxidation of resinous or oily matrices. The sealed vessel program typically ramps to 180 to 200 degrees Celsius over 10 to 15 minutes and holds for 20 to 30 minutes. The clear digest is quantitatively transferred and diluted to 25 or 50 mL with ultrapure water. Hydrochloric acid is avoided whenever the digest will be measured by ICP-MS, because chloride generates the polyatomic interference described below.
Classical alternatives include hot-plate wet digestion with nitric and perchloric or sulfuric acid mixtures, and dry ashing at 500 to 600 degrees Celsius with magnesium nitrate as a stabilization aid to prevent volatilization of arsenic during charring. These open-vessel techniques remain valid but demand stricter operator control of temperature and blank levels.
Hydride Generation Atomic Fluorescence Spectrometry (HG-AFS)
HG-AFS is the workhorse method in many Asian laboratories due to its excellent sensitivity, low running cost, and status as a national standard approach for arsenic in biological matrices. Pentavalent arsenic in the digest is first pre-reduced to the trivalent state using a mixture of potassium iodide and ascorbic acid, or thiourea combined with ascorbic acid, with a standing time of approximately 30 minutes. Arsine gas is then generated on-line by reacting the sample with an alkaline potassium borohydride solution (typically 0.5 to 2 percent KBH4 in 0.2 to 0.5 percent sodium hydroxide) in a hydrochloric acid carrier stream. The arsine is swept by argon into a heated quartz atomizer, where an argon-hydrogen flame atomizes the arsenic; excitation by an arsenic hollow cathode lamp produces fluorescence that is collected at right angles by a photomultiplier detector. Matrix separation during hydride generation largely eliminates spectral and background interferences, and instrumental detection limits below 0.05 micrograms per liter are routinely achievable.
Hydride Generation Atomic Absorption Spectrometry (HG-AAS)
HG-AAS follows the same chemical reaction pathway but quantifies arsenic by atomic absorption at the 193.7 nm resonance line in an electrically heated quartz tube atomizer. It is the classical pharmacopoeial technique for arsenic in herbal drugs and remains fully acceptable for regulatory work. Sensitivity is somewhat lower than AFS, and baseline correction with a deuterium or Zeeman background system is recommended because of the low-energy wavelength.
Inductively Coupled Plasma Mass Spectrometry (ICP-MS)
ICP-MS is the modern method of choice when simultaneous determination of arsenic, lead, cadmium, mercury, and copper is required from a single digestion. The principal analytical challenge is the polyatomic interference of argon chloride (40Ar35Cl+) on the monoisotopic arsenic signal at mass-to-charge ratio 75, which is critical for chloride-rich herbal matrices. The interference is controlled by operating in helium collision mode with kinetic energy discrimination, by avoiding hydrochloric acid during digestion, and by monitoring the chlorine proxy at mass 35 or 77 where instrument capability allows. Germanium (mass 72 or 74) is the conventional internal standard for arsenic, compensating for matrix suppression and drift. Quantitation is performed with multi-point external calibration over the range of 0 to 10 micrograms per liter, with correlation coefficients of 0.999 or better, and method detection limits in the solid sample of approximately 0.005 to 0.02 mg/kg depending on sample mass and dilution.
Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES)
ICP-OES offers high throughput and simultaneous multi-element capability, but the prominent arsenic emission lines (188.979 nm and 193.696 nm) lie in the vacuum ultraviolet region where sensitivity is limited. Practical quantitation limits in solution are generally 10 to 50 micrograms per liter, making ICP-OES suitable only for samples expected near or above the pharmacopoeial limit rather than for low-background confirmation.
Silver Diethyldithiocarbamate (Ag-DDC) Spectrophotometry
The Ag-DDC method is the historical limit test: arsine generated from the sample is absorbed in a silver diethyldithiocarbamate solution in pyridine or in a chloroform-triethylamine medium, producing a red complex measured spectrophotometrically at 510 nm. Although largely superseded by instrumental techniques, it remains a reference procedure in older monographs and requires no specialized spectrometer, though it is labor-intensive and offers only moderate precision.
Method Validation and Quality Control
Whichever technique is selected, the laboratory must demonstrate method fitness through documented validation and batch-level quality control:
- Linearity across the working calibration range with a correlation coefficient of at least 0.995, and preferably 0.999
- Spike recovery at the decision level (for example, spikes near 2 mg/kg) accepted within 80 to 120 percent, consistent with trace-element practice
- Replicate digestion precision with relative standard deviation below 10 percent, and duplicate relative percent difference below 20 percent near the limit
- Analysis of certified reference materials such as botanical leaf or tea reference matrices to verify trueness
- Procedural blanks, all-plastic acid-washed labware, high-purity reagents, and clean-bench handling to control ambient arsenic contamination
Applications
- Pharmacopoeial compliance testing and batch release of herbal materials, extracts, and finished products against the 2 mg/kg default limit and monograph-specific limits
- GMP incoming-material control, allowing manufacturers to reject contaminated crude drugs before production and to build supplier quality histories
- Export and import conformity assessment for markets applying Chinese, Japanese, United States, or Hong Kong herbal medicine standards
- Good Agricultural Practice (GAP) verification, including correlation of soil, irrigation water, and herb arsenic levels to qualify cultivation bases and remediate contaminated sites
- Product-specific risk management for formulations containing realgar or other arsenical mineral drugs, where total and speciated arsenic data support dose-based safety evaluation
- Academic and industrial research on species-dependent arsenic accumulation, processing-induced changes, extraction partitioning, and geo-authenticity studies of medicinal herbs
FAQ
What is the regulatory limit for total arsenic in Chinese herbal medicines?
The Chinese Pharmacopoeia sets a default limit of 2 mg/kg for total arsenic in herbal drugs, with several monographs specifying their own values. Export destinations may apply different limits or dose-based permitted daily exposures, so the applicable monograph and market must always be confirmed before testing.
Which method provides the best sensitivity?
ICP-MS and HG-AFS both reach sub-parts-per-billion detection in solution. ICP-MS additionally delivers simultaneous multi-element results, while HG-AFS provides comparable arsenic sensitivity at significantly lower instrument and operating cost, which is why both coexist in routine practice.
Why is chloride a concern in ICP-MS arsenic analysis?
Arsenic has only one stable isotope at mass 75, where the argon chloride polyatomic ion (40Ar35Cl+) produces a positive bias. The interference is managed by digesting without hydrochloric acid, using helium collision mode, and checking recovery with matrix-matched spikes.
How much sample is required for testing?
A representative laboratory sample of 50 to 100 g of dried material is recommended to account for heterogeneity. Each individual digestion consumes only 0.2 to 0.5 g, and at least duplicate digestions plus a spike are prepared per batch.
Does total arsenic always reflect toxicity?
Total arsenic is a conservative screening parameter. Some marine-derived or algal materials contain largely organic arsenic species of low toxicity, in which case speciation by HPLC-ICP-MS is used to quantify the inorganic fraction for accurate risk assessment.
How are results reported?
Results are reported as mg/kg on a dry weight basis, corrected using the determined moisture content, together with the method quantitation limit, recovery, and a statement of conformity against the specified limit. Results below the quantitation limit are reported as such rather than as zero.
