Atomic Emission: ICP-OES and ICP-MS

The plasma is the functional core of both instruments. Argon enters the quartz torch through three concentric tubes, the outer at roughly 15 litres per minute as coolant and the inner as the central injector that carries the sample aerosol up through the discharge. A copper coil wrapped around the top of the torch is driven by a radio-frequency generator at 27.12 MHz, typically at 1.0 to 1.5 kilowatts. A Tesla-coil spark seeds free electrons, the RF field accelerates them, they collide with neutral argon and ionise it, and the discharge becomes self-sustaining. The visible plasma sits as a teardrop above the torch with an analytically useful zone at 6000 to 7000 K and a hotter pre-heating zone reaching 10000 K.

That temperature is what distinguishes ICP from every flame source ever built. An air-acetylene flame at 2300 K can atomise sodium and potassium; a nitrous-oxide-acetylene flame at 2900 K stretches to refractory elements like aluminium and titanium but with chemical interference from oxide formation. The plasma at 6000 K and above pushes ionisation efficiency above 90 percent for elements with first ionisation energy below 8 eV (most of the periodic table) and reduces matrix-driven chemical interference to almost nothing.

Sample introduction is the other half of the source system. The default is a pneumatic concentric Meinhard nebuliser feeding a Scott double-pass spray chamber, which converts the liquid into an aerosol with droplets below 10 micrometres. High dissolved-solids samples (digested viscera, brackish groundwater) demand a Babington nebuliser whose larger orifice resists clogging at the cost of some sensitivity. For ultra-trace work an ultrasonic nebuliser raises transport efficiency by an order of magnitude.

Once the plasma has done its work, the OES arm reads the light. Excited atoms and ions relax by emitting photons at element-specific wavelengths, and the polychromator disperses them across a CCD or CID array. Modern instruments cover roughly 165 to 800 nanometres in a single read. Because the array reads all wavelengths in parallel, a 70-element panel comes off a single 30-second integration, with a practical cycle of two to three minutes per sample including rinse.

Two viewing geometries are in active use. Axial view, looking down the long axis of the plasma, gives the longest path length and roughly an order of magnitude better sensitivity, at the cost of more matrix interference. Radial view looks across the plasma at right angles to its axis, sees less matrix-driven distortion, and tolerates the high total dissolved solids that come with a typical aqua-regia digestate. Dual-view instruments switch between the two on a per-element basis, which is what FSL Madhuban runs on its Agilent 720.

The dynamic range of ICP-OES is a further practical advantage. A flame AAS calibration curve is linear over two to three orders of magnitude before self-absorption bends it; ICP-OES holds linearity over five to nine orders, so the same calibration that quantitates a part-per-billion lead level in blood also handles a part-per-thousand lead level in a paint chip without dilution.

ICP-OES addresses the multi-element panel; ICP-MS extends sensitivity to the parts-per-trillion range where ICP-OES cannot reach. The plasma at 6000 K ionises the analyte; a sampler cone with a 1 mm orifice samples the central channel into a roughing-pumped expansion region; a skimmer cone with a 0.4 mm orifice extracts the ion beam into the analyser stage. Ion lenses focus the beam, the quadrupole filter selects ions by mass-to-charge, and a discrete-dynode electron multiplier counts them. Detection limits sit between 1 and 100 parts per trillion for most metals, three to four orders below ICP-OES. That is what makes the technique mandatory for any tissue lead, blood mercury or hair arsenic question that has to discriminate ambient background from clinically relevant exposure.

The cost is a set of polyatomic interferences formed in the plasma's tail. Argon-oxide at mass 56 sits exactly on iron-56. Argon-chloride at mass 75 sits on arsenic-75 in any HCl-digested sample. Argon dimer at mass 80 wipes out selenium-80. At concentrations below 100 parts per trillion, each of these interferences is significant.

The collision and reaction cell sits between the skimmer and the quadrupole. In helium kinetic-energy-discrimination mode, the larger polyatomic ions lose more energy per collision than the monatomic analyte and are filtered out by an energy barrier at the cell exit. In reaction mode, hydrogen converts argide species back to neutral argon, and oxygen mass-shifts arsenic and selenium to their oxide forms at masses where no interference exists. Cool-plasma mode (around 600 to 800 watts) suppresses argide formation altogether at the cost of reduced ionisation for high-ionisation-energy elements. For sub-part-per-trillion accuracy in dirty matrices, BARC Mumbai's high-resolution sector ICP-MS at Trombay resolves the analyte from the polyatomic by mass alone.

Atomic absorption is a total-element technique: it quantifies lead concentration but carries no isotopic information. For some forensic questions the difference is decisive: a bullet from an encounter site in Punjab, a uranium oxide pellet found near a research facility, a smuggled cargo of saffron whose origin is contested between Iran and Kashmir. Each resolves on isotope-ratio measurement, and ICP-MS is the routine tool.

1. Lead isotope fingerprinting for bullet attribution
   Lead has four stable isotopes at masses 204, 206, 207 and 208. Three are radiogenic end-products of uranium and thorium decay, so their ratios depend on the geological age of the ore deposit. Australian Broken Hill, Missouri Mississippi Valley and Indian Rajpura-Dariba leads each carry a distinct 206/204 and 208/206 fingerprint preserved through smelting and bullet manufacture. Quadrupole ICP-MS reads the ratios to roughly 0.1 percent precision; a multicollector sector refines this to a few parts per million for smelter-level attribution.
2. Mercury isotopes for methyl versus inorganic exposure
   Mercury has seven stable isotopes whose mass-dependent and mass-independent fractionation patterns differ between dietary methylmercury (from fish and rice) and industrial inorganic mercury (cinnabar, dental amalgam, chlor-alkali effluent). Hair-segment ICP-MS work on a chronic exposure case can separate the two pathways via delta-202 and capital-delta-199 values, increasingly accepted in environmental-health litigation.
3. Strontium-87 over strontium-86 for geographic provenance
   The 87Sr/86Sr ratio in soil and water reflects the age and rubidium content of the underlying bedrock, and passes unchanged through the food web into bone, hair and tooth enamel. Used in Europe to determine the geographic origin of unidentified bodies and in food-fraud cases for verifying the regional provenance of saffron, basmati rice and Darjeeling tea. NIPER Mohali and IIT Bombay have multicollector capability for the highest-precision work.
4. Uranium-235 over uranium-238 for nuclear forensics
   Natural uranium runs at 0.72 percent. Reactor-grade enriched uranium runs at 3 to 5 percent, weapons-usable highly enriched uranium above 20 percent, depleted uranium around 0.2 percent. ICP-MS measurement of the 235/238 ratio is the first analytical question asked of any unattributed uranium sample, and BARC Trombay is the operational facility in India under the Atomic Energy Act.

Isotope-ratio precision on a quadrupole is bounded by counting statistics of the rarer isotope. A multicollector sector measures all isotopes simultaneously on separate Faraday cups and gets to part-per-million precision, but the capital cost is an order of magnitude higher. The practical Indian division: quadrupole ICP-MS at the CFSL bench for routine attribution work, sector instruments at BARC for the small fraction of cases where precision matters more than throughput.

The facilities are not interchangeable. A toxic-metal viscera submission from a Haryana case and a uranium-attribution submission from a Mumbai customs seizure both require ICP-MS, but they route to entirely different laboratories with different instrument configurations, calibration regimes, and analyst pools.

• CFSL Chandigarh. Agilent 7700 ICP-MS with the high-matrix-introduction interface and helium collision-cell mode. Standard panel covers arsenic, lead, mercury, cadmium, thallium, antimony, chromium and manganese in viscera, blood, urine, hair and nail. Calibration uses external standards bracketed with rhodium, indium and terbium internal standards spanning low, mid and high mass.
• FSL Madhuban Sector 14 (Haryana). Agilent 720 ICP-OES, dual-view, handling the trace-metal volume from the National Capital Region toxicology overflow and the Haryana state caseload. The OES rather than MS choice was deliberate: throughput is higher, and most criminal casework involves lethal-range concentrations where parts-per-billion sensitivity is sufficient. Cases that need parts-per-trillion work route up to CFSL Chandigarh.
• WBHIDCO and the West Bengal groundwater programme. ICP-MS deployed for the long-running arsenic surveillance work across the Bengal delta, where parts-per-trillion sensitivity and helium collision mode are mandatory because of the chloride-rich matrix.
• NEERI Nagpur. Multiple ICP-MS instruments for environmental forensic work on industrial discharges, soil contamination and ambient particulate metal loading.
• BARC Trombay (Mumbai). Multicollector high-resolution ICP-MS for nuclear-material attribution under the Atomic Energy Act and the IAEA safeguards agreement.

QC on a routine ICP-MS shift looks similar across the better Indian labs: a reagent blank, a calibration curve covering the working range, an internal standard in every vial to correct drift and matrix suppression, a continuing calibration verification standard every 10 samples, a duplicate every 20, and a CRM at the start, middle and end of each batch. Any failure of the 10-sample QC check requires reanalysis back to the previous passing QC, which is the structural reason ICP-MS reports take days rather than hours despite short instrument time per sample.

Forensic applications run wider than the toxic-metals panel. Bullet lead profiling combines the 204/206/207/208 isotope pattern with a trace-element fingerprint of antimony, arsenic, copper and tin. Glass elemental fingerprinting on a microsampled fragment digested in HF and HNO3 matches shards from a hit-and-run windshield to a suspect vehicle through rare-earth element abundances. Gunshot residue work on hand swabs uses the barium-antimony-lead diagnostic triad. Heavy-metal contamination of Ayurvedic bhasmas (lead and mercury at percent-level concentrations in commercially marketed preparations) is contracted by FSSAI to NABL-accredited ICP-MS work for prosecution.