Atomic Emission Spectroscopy: Identifying Elements by Their Light

Every element glows with its own set of colours. Heat sodium and you get a strong yellow; lithium gives crimson; copper gives a blue-green. A school flame test uses this idea to identify metals by eye. Atomic emission spectroscopy (AES) takes the same idea and turns it into one of the most powerful tools in analytical chemistry, capable of measuring dozens of elements at once down to parts per billion.

This article explains the physics behind the technique and then how real instruments use it.

The physics: excited atoms give out light

Atoms have fixed energy levels. In the ground state, electrons sit in the lowest levels available. Give the atom enough energy, by heating it in a flame or plasma, and an electron can jump to a higher level. The atom is now in an excited state.

Excited states don’t last long, typically around ten nanoseconds. The electron falls back down and the atom releases the energy difference as a photon:

E_photon = hν = hc ⁄ λ = E_upper − E_lower

Because the energy levels are fixed for each element, the photons come out only at particular wavelengths. The result is a line spectrum, a set of sharp, element-specific wavelengths. Sodium’s strongest emission is the pair of lines near 589 nm, which is why sodium street lamps look so yellow. Potassium’s strongest visible lines lie in the deep red near 766 nm, and lithium’s at 671 nm.

Two facts make this useful for analysis:

1. Which wavelengths appear tells you which elements are present (qualitative analysis).
2. How bright each line is tells you how much of the element is present (quantitative analysis). More atoms emitting means more photons at that wavelength.

The general layout of an emission instrument

Every atomic emission instrument has the same basic parts:

1. Sample introduction. Most samples are dissolved in dilute acid. A nebuliser sprays the solution into a fine mist, and a spray chamber removes the larger droplets.
2. Atomisation and excitation source. Something hot enough to evaporate the droplets, break compounds into free atoms and excite those atoms: a flame, an electric arc or spark, or a plasma.
3. Wavelength selector. A diffraction grating (monochromator or polychromator) separates the light into its wavelengths.
4. Detector. A photomultiplier tube or, in modern instruments, a solid-state array detector measures the intensity at each wavelength.
5. Data system. Software converts intensities into concentrations using calibration standards.

The biggest difference between types of AES is the excitation source.

Flame photometry: the simple version

The oldest and cheapest form uses an ordinary flame, often air–propane or air–acetylene, at roughly 1,700 to 2,300 °C. That temperature is enough to excite elements with low-lying energy levels, mainly the alkali and alkaline earth metals.

A flame photometer typically uses simple optical filters rather than a full spectrometer to pick out one element’s light, for example the sodium line at 589 nm or the potassium line at 766 nm. It is still widely used for:

• sodium and potassium in blood serum and urine in some clinical labs;
• potassium in fertilisers and soils;
• sodium in foods and drinks.

Its limitation is temperature. A flame simply isn’t hot enough to excite most transition metals or non-metals effectively.

ICP-OES: the modern workhorse

The dominant form of atomic emission today is inductively coupled plasma optical emission spectroscopy (ICP-OES, also called ICP-AES).

A stream of argon gas flows through a quartz torch surrounded by a copper coil carrying a radio-frequency current. A spark seeds the argon with electrons, and the oscillating magnetic field accelerates them, ionising more argon. The result is a stable, brilliant plasma with temperatures of roughly 6,000 to 10,000 K, hotter than the surface of the Sun.

At those temperatures:

• compounds are broken down almost completely into atoms;
• most elements are excited, and many are ionised, so both atomic and ionic lines appear;
• about 70 elements can be measured, often simultaneously.

A modern ICP-OES can analyse a water sample for 30 elements in a couple of minutes, with detection limits for many metals in the range of micrograms per litre. For even lower limits, laboratories use the related ICP-MS, which counts ions by mass rather than measuring light. The two are compared in AAS vs ICP.

How concentrations are measured

Emission intensity depends on instrument settings, so AES always needs calibration:

1. Prepare standard solutions of known concentration, for example 0, 1, 5, 10 and 20 mg L⁻¹ of the element.
2. Measure the emission intensity of each at the chosen wavelength.
3. Plot intensity against concentration to make a calibration curve.
4. Measure the unknown and read its concentration from the line.

Over a useful range, intensity is proportional to concentration. At high concentrations the curve bends because of self-absorption: cooler atoms at the edge of the flame or plasma absorb some of the light emitted from the hot centre. Diluting the sample restores linearity.

An internal standard (an element added at a fixed concentration to every solution, such as yttrium) is often used to correct for drift and for differences in how easily solutions are sprayed.

Interferences to watch for

No technique is perfect. The main problems in AES are:

• Spectral interference. Two elements emit at nearly the same wavelength, or a line sits on a bright background. The fix is to choose a different line, use higher-resolution optics or apply background correction.
• Chemical interference (mainly flames). The element forms a compound that doesn’t break apart in the flame. Phosphate, for instance, suppresses calcium emission in cooler flames by forming stable calcium phosphate species. Hotter sources like the ICP largely avoid this.
• Ionisation interference. In a hot source, some atoms lose an electron and stop emitting their atomic lines. Adding an easily ionised element (such as caesium) as an “ionisation buffer” can stabilise the signal in flame work.
• Matrix effects. High salt or acid content changes how the sample is nebulised and how it behaves in the plasma. Matching the standards to the sample matrix, or using internal standards, compensates.

Where atomic emission spectroscopy is used

• Environmental monitoring: heavy metals such as lead, cadmium and arsenic in drinking water, rivers and soils (see water testing).
• Food and agriculture: mineral nutrients such as calcium, magnesium, iron and zinc in foods, feeds and plant tissue.
• Metallurgy: checking the composition of alloys, often by spark OES directly on solid metal.
• Geology and mining: major and trace elements in rocks and ores.
• Petroleum: wear metals in engine oil, which reveal which engine parts are wearing out.
• Pharmaceuticals: limits on elemental impurities in medicines.

Emission vs absorption: how AES relates to AAS

Atomic absorption spectroscopy uses the same energy levels but looks at the opposite process. Instead of measuring light given out by excited atoms, it shines light at the element’s wavelength and measures how much ground-state atoms absorb.

• AES measures emission; AAS measures absorption.
• AES (especially ICP-OES) handles many elements at once; classic AAS measures one element at a time.
• AAS instruments are cheaper and very reliable for a single element; ICP-OES is faster for large element panels.

Common misconceptions

• “Each element has one colour.” Each element has many emission lines. The flame colour you see is just the mix of the strongest visible lines.
• “Brighter always means more.” Only under calibrated, consistent conditions, and only within the linear range.
• “It measures compounds.” AES measures total element content. Iron from rust and iron from a supplement give the same iron lines.

Key takeaways

• Heated atoms emit light at element-specific wavelengths when excited electrons fall back to lower energy levels.
• The wavelength identifies the element; the intensity, after calibration, gives its concentration.
• Flame photometry is simple and cheap but mostly limited to alkali and alkaline earth metals.
• ICP-OES uses an argon plasma at 6,000 to 10,000 K to measure about 70 elements quickly and at trace levels.
• Calibration curves, internal standards and careful choice of lines deal with most interferences.
• For the everyday version of the same physics, try the classic flame test.