ICP-MS and ICP-OES: elemental analysis with an inductively coupled plasma

An ICP-MS instrument has six parts: a sample introduction system, the inductively coupled plasma that acts as ion source, an interface, electrostatic lenses, a mass spectrometer and a detector (Mazarakioti et al. 2022). A peristaltic pump feeds the liquid sample to a nebuliser, where it is mixed with argon into an aerosol; the spray chamber removes large droplets, and the fine aerosol passes through the central channel of the argon plasma (Mazarakioti et al. 2022). There the droplets are vaporised, decomposed, atomised and finally ionised. Argon has a higher ionisation potential than most elements, so they are ionised efficiently, mainly to singly charged positive ions (Mazarakioti et al. 2022).

The ions pass through the ~1 mm orifice of the sample cone and the ~0.45 mm orifice of the skimmer cone, with 150–300 Pa between them, into the high vacuum where ion lenses steer them to the mass analyser and away from photons and neutral species (Mazarakioti et al. 2022). A quadrupole (or triple quadrupole, time-of-flight or sector-field analyser) separates them by mass-to-charge ratio, and an electron multiplier counts them (Mazarakioti et al. 2022; Gregar et al. 2025). Counts are converted to concentrations with calibration curves, and internal-standard elements of similar mass and ionisation potential correct for drift and matrix effects (Mazarakioti et al. 2022).

ICP-OES uses an inductively coupled plasma too but measures the light that excited atoms and ions emit at element-specific spectral lines, for example the Hg lines at 253 and 194 nm, instead of counting ions (Provete et al. 2024). Its sensitivity is normally lower than that of ICP-MS (Provete et al. 2024), but it is a standard way to determine major constituents such as lithium in cathode materials (Recknagel et al. 2025).

• Calibration curve. For each element the signal (counts per second for ICP-MS) is plotted against standards of known concentration, and sample concentrations are read from it (Mazarakioti et al. 2022).
• Internal standard recovery. If the analyte-to-internal-standard ratio stays within about ±20% recovery, drift and matrix effects are under control (Mazarakioti et al. 2022).
• From solution to solid. Results for a digested solid are converted back with the sample mass and dilution volume; for marine sediment a 100-fold dilution turned an instrumental Hg LoQ of 0.026 µg L⁻¹ into a method LoQ of 1.9 µg kg⁻¹ (Provete et al. 2024).
• Metal loading. Results are usually reported as wt% and molar ratios: ICP-OES of Pt-doped NiFe MOFs gave 0.52, 1.05, 1.57 and 2.08 wt% Pt for nominal 0.5–2.0 wt% and a Ni:Fe ratio of about 3:2 in every sample (Bekmyrza et al. 2025).
• Isotope ratios. ICP-MS also measures isotopic abundances (Mazarakioti et al. 2022); a multicollector instrument gave a lithium isotope ratio of 12.34 for an NMC reference cathode (Recknagel et al. 2025).
• Single-particle data. In spICP-MS the time trace shows discrete pulses above a continuous ionic background; a statistical threshold separates particle events, pulse height is converted to particle mass and size, and the pulse rate gives particle number concentration (Gregar et al. 2025).

• Digest the solid. ICP was designed for liquids, so solids are normally dissolved by thermally assisted acid digestion, commonly with 65–70% HNO₃ in a microwave oven, or with HNO₃ plus H₂O₂, HCl or HF (Mazarakioti et al. 2022). A typical microwave programme ramps to 180 °C in 5 min and holds for 10 min (Provete et al. 2024).
• Aqua regia for metals and oxides. NMC cathode powder dissolves in aqua regia on a hotplate at 120–150 °C, and some laboratories add HF (Recknagel et al. 2025); Pt-doped NiFe MOFs were digested in aqua regia (HCl:HNO₃ 3:1) at 80 °C for 12 h (Bekmyrza et al. 2025).
• Dilute and add internal standards. Make up to volume with ultrapure water, add internal standards such as Sc or In, and match the calibration matrix when measuring traces next to large amounts of Li, Ni, Mn and Co (Mazarakioti et al. 2022; Recknagel et al. 2025).
• Weigh enough. Sample intakes of 20–500 mg were used for major elements of NMC and 50–1000 mg for traces (Recknagel et al. 2025).
• Nanoparticle suspensions. For spICP-MS, dilute until individual particles arrive one at a time, and calibrate transport efficiency, usually with gold nanoparticle standards (Gregar et al. 2025).
• Protect reactive powders. NMC 111 takes up CO₂ from air (its carbon content rose from 0.066 to 0.094% in 5 months), which changes the mass fractions, so it is stored under argon (Recknagel et al. 2025).

Composition and metal loading. ICP confirms that a synthesis gave the intended stoichiometry or dopant level, as in Pt-doped NiFe metal–organic frameworks whose measured Pt contents matched the nominal loadings (Bekmyrza et al. 2025). The same digest-and-measure approach applies to frameworks such as HKUST-1, ZIF-8 and UiO-66.

Battery materials. Elemental analysis of cathodes is critical for quality, performance and recycling; manufacturers hold Li, Ni, Mn and Co ratios within a narrow margin and limit impurities such as Fe, Na and Cr to the mg kg⁻¹ range, and the first certified reference material for an NMC 111 cathode was characterised mainly by ICP-OES and ICP-MS (Recknagel et al. 2025). See NMC, LiFePO₄ and our overview of battery materials.

Catalyst stability. Metal dissolved into the electrolyte during electrolysis, measured by ICP-MS, gives the stability number (molecules of product per dissolved metal atom) and the activity–stability factor of water-splitting catalysts (Sharma et al. 2026); pair it with cyclic voltammetry and EIS.

Nanoparticles. Single-particle ICP-MS gives element-specific core size, size distribution and particle number concentration, complementing TEM, SEM and DLS (Gregar et al. 2025). In the Pt-doped MOF study, ICP-OES gave the composition while XPS probed the electronic structure of Pt, Ni and Fe (Bekmyrza et al. 2025). All are covered in our guide to characterisation techniques.

• Spectroscopic interferences. Isobaric isotopes of other elements, doubly charged ions, polyatomic ions and tailing from neighbouring peaks can sit at the analyte's mass (Mazarakioti et al. 2022). Reaction gases such as H₂, O₂, NH₃ or CH₃F in tandem ICP-MS/MS remove many of them (Balcaen et al. 2015).
• Matrix effects. Sample introduction, plasma and space-charge effects suppress or enhance signals, and dissolved salts gradually deposit on the cones and narrow their orifices (Mazarakioti et al. 2022). Use internal standards and matrix-matched or standard-addition calibration.
• Forgetting the dilution. Digestion and dilution raise the method detection limit well above the instrumental one (Provete et al. 2024).
• Losses and memory. Hg can be lost during digestion or stick in the introduction system, and high standards leave a memory effect that needs long rinses (Provete et al. 2024).
• spICP-MS assumptions. Sizes assume a known composition, density and spherical shape; quadrupole instruments see one element at a time; and dissolved ions raise the background, which limits detection of particles below roughly 10 nm; for Ti and Si the smallest detectable particles in biological samples were as large as 48.6 and 350 nm (Gregar et al. 2025).
• Cost. ICP-MS and ICP-OES are expensive instruments that consume large volumes of argon and need sample treatment (Provete et al. 2024).

The LabCanvas figure editor Library has the ICP-MS instrument as an editable object, including a cutaway of the nebuliser, plasma torch, sample and skimmer cones and the ion path to the quadrupole and detector, plus an editable data plot for calibration curves, mass spectra or single-particle time traces. Label the element and isotope on each peak, show the digestion step as a small schematic, and export the panel for a methods figure, SI or graphical abstract.