Atomic Absorption Spectroscopy (AAS)

Why AAS matters

• Trace metals at clinical and regulatory thresholds. Lead in blood (CDC reference 3.5 µg/dL = 35 µg/L), cadmium in drinking water (EPA MCL 5 µg/L), mercury in fish (FDA action level 1 mg/kg) all demand quantitation 100x below action limits — a window where GFAAS at 0.01 µg/L excels.
• Element-specific selectivity. Each hollow-cathode lamp emits only its element's resonance line; the monochromator only has to isolate that line from neighbors, not perform high-resolution scanning. Spectral interferences from other elements are rare because the chance of overlap within ±0.001 nm is small.
• Cheap and rugged for single-element work. A bench AAS costs roughly USD 30,000 to 80,000, an order of magnitude below ICP-MS. Operating costs are low: acetylene gas, electricity, and a USD 800 lamp per element. For a clinical lab running blood lead 200 times a day, GFAAS dollars-per-result is unbeatable.
• Mature regulatory acceptance. EPA Methods 7000B (FAAS) and 7010 (GFAAS), AOAC, ASTM D3559 (lead in water by AAS), USP and Ph. Eur. monographs all reference AAS as a primary method. Audit trail and reference materials are abundant.
• Hydride and cold-vapor accessory. Hg, As, Sb, Bi, Se, Te volatilize as hydrides (NaBH4 reduction) or as elemental vapor (SnCl2 reduction for Hg) and bypass the flame/furnace, gaining 10 to 100x sensitivity. Cold-vapor AAS for mercury reaches 0.01 µg/L without GFAAS hardware.
• Specific to atomic populations, not molecules. Sample digestion converts everything (organic Pb, inorganic Pb²⁺, Pb-protein) to free Pb atoms in the atomizer, so AAS reports total elemental concentration. Speciation requires upstream separation (HPLC-AAS coupling).
• Foundation for most metals analysis training. AAS is still the introductory trace-metal technique in analytical chemistry programs because the optical principle is identical to UV-Vis (Beer-Lambert) and the atomization chemistry illustrates flame thermodynamics tangibly.

Common misconceptions

• AAS measures the metal as it is in the sample. No — atomization destroys all chemistry. Free Pb atoms in the flame come equally from Pb²⁺ nitrate and tetraethyl-Pb. AAS is total-element only; chemical-form information is gone after atomization.
• Higher temperature is always better. Refractory elements (Al, Si, Mo, V, Ti) do need the 2900 °C nitrous-oxide–acetylene flame, but volatile elements (Na, K, Pb, Cd) are partially ionized at high temperature, which removes them from the absorbing atomic state and crashes sensitivity. Use the coolest flame that still atomizes.
• Calibration in pure water is fine. Real samples carry matrix elements that affect ionization, atomization efficiency, and viscosity (which affects nebulizer uptake). Use matrix-matched standards or, when matrix is unknown, the method of standard additions.
• Background correction handles all interferences. Deuterium-lamp correction handles broad-band background up to ~0.5 absorbance, but cannot correct sharp molecular bands (e.g. PO4 at 213 nm under Zn). Zeeman correction handles structured backgrounds; without it, results in biological matrices may be biased high by 20 to 200%.
• One method works across the dynamic range. AAS has a ~3 decade linear range. Below LOQ you need GFAAS or pre-concentration; above ~5 absorbance you must dilute or use a less-sensitive line (e.g. 217.0 nm Pb instead of 283.3 nm — ε is 3x lower).
• The hollow-cathode lamp is forever. Lamp lifetime is roughly 5,000 mA-hours; sputter coating accumulates inside the envelope and dims the line. Single-element lamps last 3 to 5 years in routine use; multi-element lamps less. Budget 4 to 8 lamp replacements per year for a busy lab.

How an AAS measurement is made

Calibration first. Aspirate a 0 µg/L blank, then 5 standards spanning the expected sample range (e.g., 0.5, 1, 2, 5, 10 µg/L for blood lead in GFAAS). Each standard occupies the atomizer for a fixed program: in flame, that is a steady aspiration giving a flat absorbance for 5 seconds; in graphite furnace, a temperature ramp dries the 20 µL drop at 110 °C, chars off matrix at 600 to 1200 °C, atomizes at 1800 to 2400 °C in a 1 to 2 second pulse, and cleans at 2600 °C. The atomization step records a peak absorbance; integrated absorbance (peak area) is preferred because it is robust to atomization-rate variation. Plot peak area or peak height versus concentration; least-squares fit gives the slope k. Run the unknown identically and read concentration from the curve. Measure a check standard every 10 samples to detect drift, and re-calibrate if the check is off by more than 10%.

Sample preparation is the make-or-break step. Solids must be digested into homogeneous solution (concentrated HNO3 or HNO3/H2O2 in a microwave digester), and biological matrices (whole blood, urine, serum) are diluted 1:5 to 1:50 with 0.5% Triton X-100 or 0.2% nitric acid. Matrix modifiers — Pd(NO3)2 + Mg(NO3)2 for Pb, Cd; NH4H2PO4 for Cd; ascorbic acid for Hg — co-deposit with the analyte to allow higher char temperatures without analyte loss. The modifier-element pairing is published in the GFAAS standard methods (EPA, ISO 15586) and is critical for reaching nominal LODs. Without modifiers, char temperatures must stay below the volatilization point and matrix removal is incomplete, raising background and ruining detection limits.

AAS vs ICP-OES vs ICP-MS vs XRF — choosing the right metals technique

Property	FAAS	GFAAS	ICP-OES	ICP-MS	XRF	HG-AAS
Detection limit	1–100 µg/L (ppb)	0.01–1 µg/L (ppt)	1–10 µg/L	0.001–0.1 µg/L (ppt)	10–100 mg/kg	0.01–0.1 µg/L
Multi-element	Sequential	Sequential, slow	Simultaneous, ~30 elements	Simultaneous, ~75 elements	Simultaneous	Sequential
Sample form	Solution	Solution, micro-volume	Solution	Solution	Solid (direct)	Solution after reduction
Sample throughput	60–120 / hour	20–40 / hour	30+ elements in 30 s	30+ elements in 60 s	1–10 min, no prep	30–60 / hour
Approximate cost (USD)	30–50 k	60–100 k	100–200 k	200–500 k	50–250 k (XRF)	40–80 k
Argon consumption	None	~1 L/min Ar purge	15–18 L/min	15–20 L/min	None	None
Isotope information	No	No	No	Yes	No	No
Best for	Routine 1–10 elements	Trace clinical, single-element	Routine 30 element scans	Ultra-trace, isotopes	Alloys, ores, in-situ	As, Sb, Se, Hg, Bi

Famous applications and detection floors

• CDC blood lead surveillance. The reference value of 3.5 µg/dL (35 µg/L) is roughly 7000x above the GFAAS detection limit of 0.05 µg/L. CDC's 2017 NHANES data set (used by the WHO Lead Working Group) was generated almost entirely on GFAAS-Zeeman instruments at state public-health laboratories — about 8 million blood lead measurements were compiled this way before ICP-MS supplanted GFAAS in the largest reference labs.
• EPA drinking-water lead and copper rule. 15 µg/L Pb action level, 1300 µg/L Cu action level. Method 200.9 (GFAAS) provides LOD ≈ 0.5 µg/L Pb and was the basis for early Flint, Michigan crisis reporting (2014–2016) before ICP-MS Method 200.8 became dominant.
• Iron-deficiency anemia screening. Serum iron is measured by FAAS at 248.3 nm with detection ≈ 5 µg/L; combined with total iron-binding capacity, it underpins the WHO/UNICEF anemia diagnostic protocol used in 100+ low- and middle-income country health surveys.
• Alan Walsh's foundational paper. Walsh's 1955 Spectrochimica Acta paper introduced the hollow-cathode lamp / flame atomizer combination and predicted detection limits in the 0.01 ppm range. CSIRO commercialized the technology through Hilger & Watts in 1962, and AAS displaced classical wet-chemistry titrations for trace metals across the 1960s–70s.
• Mercury in environmental water — cold vapor AAS. EPA Method 245.1 reduces Hg²⁺ to Hg(0) with SnCl2, sweeps it into a quartz absorbance cell at room temperature; LOD ≈ 0.05 µg/L. The Minamata Convention's compliance monitoring relies on cold-vapor AAS at most signatory laboratories.