A Researcher's Guide to Characterising Nanomaterials: Key Techniques and What They Tell You

Why Characterisation Is Non-Negotiable

A nanomaterial that is not well characterised is a variable, not a material. The properties that make nanomaterials interesting — quantum confinement effects, surface plasmon resonance, catalytic activity, biological interaction — are all acutely sensitive to size, shape, surface chemistry, and purity. Two batches of nominally identical "20 nm gold nanoparticles" from different suppliers (or different synthesis runs) may have peak SPR wavelengths 15 nm apart, size distributions three times as broad, and vastly different colloidal stabilities.

Rigorous characterisation protects reproducibility, enables troubleshooting when experiments fail, and is increasingly required for publication in high-impact journals and for regulatory submissions in applied contexts. This guide covers the core analytical toolkit.

Electron Microscopy: Seeing Individual Nanoparticles

Transmission Electron Microscopy (TEM)

What it measures: Direct imaging of individual nanoparticles at atomic resolution in some cases. Provides size, shape, internal structure (crystallinity, core-shell geometry, defects), and size distribution from statistical analysis of hundreds of particles.

Sample preparation: Dilute dispersion deposited on a carbon-coated TEM grid and dried. For biological samples, negative staining or cryo-TEM preparation.

Key specifications it validates: Stated particle diameter, monodispersity, aspect ratio of nanorods/tubes, shell thickness for core-shell structures, number of walls in MWCNTs.

Limitations: Measures particles in the dry state — colloidal behaviour in solution can differ. Statistical sampling requires manual measurement of many particles; automated analysis software helps but introduces artefacts. High-resolution TEM requires specialist operation.

Variants: STEM (scanning TEM) for elemental mapping via EDX; HRTEM for lattice-resolution imaging; cryo-TEM for hydrated samples.

In depth: transmission electron microscopy for nanomaterials — grid preparation, imaging modes, and how to build a size distribution that survives peer review.

Scanning Electron Microscopy (SEM)

What it measures: Surface morphology and topology of nanoparticles and nanostructured surfaces. Lower resolution than TEM (~1–5 nm), but larger field of view and easier sample preparation for solid samples.

Best for: Nanoparticle films, CNT mats and forests, nanostructured surfaces, catalyst supports, porous materials. Less suitable for isolated sub-20 nm particles.

Variants: FE-SEM (field emission SEM) for higher resolution; SEM-EDX for elemental analysis.

In depth: scanning electron microscopy of nanomaterials — detector modes, charging and coating, and what a working distance actually buys you.

Dynamic Light Scattering (DLS): Hydrodynamic Size in Solution

What it measures: The hydrodynamic diameter — the effective size of the nanoparticle including its surface layer (stabilising ligands, protein corona in biological media, solvent layer) as it diffuses in solution. Derived from the Stokes-Einstein equation using measured diffusion coefficients.

Why it differs from TEM size: DLS measures the hydrodynamic particle, which is larger than the core. A 10 nm gold nanoparticle with a PEG-2000 coating may show a DLS size of 25–30 nm. A citrate-capped particle measured at 15 nm by TEM may show 18–22 nm by DLS. These differences are expected and informative.

Key metric: Polydispersity index (PDI) — values below 0.1 indicate narrow, monodisperse distributions; above 0.3 indicates significant polydispersity.

Also provides: Zeta potential (electrokinetic potential at the particle surface, measured by electrophoretic light scattering). Zeta potential is the key indicator of colloidal stability: |ζ| > 30 mV indicates electrostatically stabilised colloid; values near 0 suggest imminent aggregation.

Limitations: DLS reports intensity-weighted distributions that overweight large particles; a few aggregates can dominate the measurement. Not suitable for polydisperse samples without complementary TEM. Minimum particle size ~1 nm.

In depth: DLS and zeta potential, in full — reading a correlogram, when the PDI is lying to you, and what a zeta potential does and does not predict about shelf life.

X-Ray Diffraction (XRD): Crystal Structure and Crystallite Size

What it measures: The crystallographic phase of the material (confirms you have what you think you have), lattice parameters, and crystallite size via the Scherrer equation applied to peak broadening.

Key applications:

• Confirming the crystal phase of metal oxide nanoparticles (anatase vs rutile TiO₂, magnetite vs maghemite iron oxide)
• Calculating crystallite size from peak width — peaks broaden as crystallite size decreases below ~100 nm
• Measuring degree of graphitisation in carbon nanomaterials (graphite d₀₀₂ spacing, G-band position)
• Confirming alloy composition in bimetallic nanoparticles

Limitations: Reports the crystallite size (coherent diffraction domain), which may differ from particle size — nanoparticles may be polycrystalline or have amorphous shells. Requires relatively pure, dry powder samples. Poor sensitivity for surface species.

In depth: XRD phase analysis and crystallite sizing — the Scherrer equation and where it breaks down, ICDD phase matching, and Rietveld refinement on nanoscale powders.

Raman Spectroscopy: Vibrational Fingerprinting for Carbon Nanomaterials

What it measures: Inelastic scattering of photons from vibrational modes, producing a molecular fingerprint. Particularly powerful for carbon nanomaterials where specific Raman peaks are diagnostic.

For graphene, Raman spectroscopy provides:

• D band (~1350 cm⁻¹) — defect density indicator; high D/G ratio indicates structural disorder
• G band (~1580 cm⁻¹) — sp² carbon stretching; present in all graphitic materials
• 2D band (~2700 cm⁻¹) — highly sensitive to layer number; monolayer graphene shows a sharp, symmetric 2D peak; bilayer/few-layer shows a broader, split 2D band

For carbon nanotubes, Raman provides:

• Radial breathing mode (RBM, 100–350 cm⁻¹) — frequency inversely proportional to CNT diameter; presence confirms nanotube structure
• D/G ratio — defect and purity assessment; pure SWCNTs have very low D/G ratios
• Electronic type enrichment — metallic vs semiconducting SWCNT contributions at different laser energies

For other nanomaterials: Phase identification of TiO₂ (anatase/rutile peaks), characterisation of semiconductor QDs, identification of surface-enhanced species in SERS.

In depth: reading the D, G and 2D bands — what the D-to-G intensity ratio does and does not tell you about defect density, and how the 2D band counts graphene layers.

X-Ray Photoelectron Spectroscopy (XPS): Surface Elemental and Chemical State Analysis

What it measures: Elemental composition and chemical bonding state of the top 5–10 nm of a surface. Provides atom percent of each element and, crucially, distinguishes chemical states (e.g. metallic Au⁰ vs ionic Au³⁺; C-C vs C-O vs C=O in graphene oxide).

Key applications in nanomaterials:

• Quantifying C:O ratio in graphene oxide and reduced graphene oxide — confirms reduction success
• Confirming surface ligand attachment chemistry (e.g. thiol-Au bonding in functionalised gold nanoparticles)
• Measuring metal oxidation states in oxide nanoparticles
• Confirming doping (e.g. nitrogen doping of graphene via N 1s peak)
• Catalyst characterisation — active metal speciation on support

Limitations: Surface-only technique (5–10 nm depth); bulk composition may differ from surface. Requires UHV conditions; samples must be compatible. Relatively slow measurement; specialist instrument.

In depth: XPS surface chemistry and oxidation state — binding energy shifts, charge referencing, and why a 10 nm sampling depth changes what a composition number means.

BET Surface Area Analysis

What it measures: Total specific surface area (m²/g) by gas adsorption isotherm analysis (Brunauer-Emmett-Teller theory). Also provides pore size distribution for porous nanomaterials.

Most relevant for: High-surface-area materials where surface area drives performance — catalysts, adsorbents, electrode materials, activated carbon, mesoporous silica, MOFs, graphene nanoplatelets.

Interpretation: Compare measured BET surface area to the theoretical surface area calculated from TEM-derived particle size assuming spherical, non-porous particles. Significantly lower BET SA than theoretical indicates aggregation or sintering; significantly higher indicates porosity or surface roughness.

In depth: what a BET number actually means — choosing the linear range, degassing that does not sinter the sample, and converting area per gram into an equivalent diameter.

Thermogravimetric Analysis (TGA)

What it measures: Mass change as a function of temperature in controlled atmosphere. Used for CNT and graphene purity assessment (carbon content), organic ligand loading on nanoparticle surfaces, and thermal stability characterisation.

For CNTs: Combustion temperature in air indicates wall quality (defective MWCNTs combust at lower temperatures than high-quality SWCNTs). Residual mass at high temperature = inorganic catalyst content.

In depth: TGA for purity, loading and thermal stability — separating ligand loss from decomposition, reading residual mass as catalyst content, and why the atmosphere decides the answer.

Which technique answers which question

The eight techniques above overlap, and the overlaps are where most wasted instrument time goes. This is the short version: what each one actually measures, over what size range, and what it costs you to find out.

Technique	What it measures	Size range	Sample state	Destructive?	Time per sample
TEM	Projected size, shape, internal and lattice structure, one particle at a time	~0.1 nm to 1 um	Dry, on a grid; or vitrified for cryo	No, but the sample is consumed as prepared	Hours, including counting enough particles for statistics
SEM	Surface morphology and topology, and elemental maps with EDX	~1 nm to millimetres	Dry solid, often sputter-coated	No	Under an hour for a survey
DLS	Hydrodynamic diameter and polydispersity; zeta potential on the same instrument	~1 nm to 1 um	Dilute suspension	No; the sample is recoverable	Minutes
XRD	Crystalline phase, lattice parameters, crystallite size from peak broadening	Crystallites below ~100 nm are where broadening is usable	Dry powder, or a film in grazing incidence	No, but tens to hundreds of mg are tied up	Half an hour to several hours, depending on the scan
Raman	Vibrational fingerprint; defect density and layer number in carbon materials	Not a sizing method; probes a spot of roughly a micrometre	Powder, film, suspension or in situ	No, unless the laser burns the sample	Minutes per spot
XPS	Elemental composition and chemical state of the top 5 to 10 nm	Surface only, at any particle size	Dry solid, ultra-high vacuum compatible	No, but the sample cannot come back wet	Hours, plus queue time on a shared instrument
BET	Specific surface area, and pore size distribution for porous solids	Reports area per gram rather than a size	Dry powder, degassed first	No	Hours, most of it degassing
TGA	Mass loss against temperature: ligand loading, carbon purity, residual catalyst	Not a sizing method	Dry powder, tens of mg	Yes — the sample is burned	An hour or two per ramp

Two rules follow from the table and are worth stating plainly. First, no single row answers "how big is it" on its own: TEM measures the dry particle, DLS measures the particle plus everything solvated around it, and XRD measures the crystallite inside it. Three different numbers for one material, all correct. Second, only TGA destroys the sample — so run it last.

Putting It Together: A Minimum Characterisation Set

For most published nanomaterial research, reviewers expect at minimum:

1. Morphology and size — TEM with size statistics
2. Hydrodynamic size and zeta potential in working medium — DLS
3. Crystal structure / phase confirmation — XRD or Raman (material-dependent)
4. Surface chemistry — XPS or FTIR for functionalised materials
5. Purity — TGA for carbon nanomaterials; ICP-OES for metal contaminants

When sourcing nanomaterials from suppliers — whether on nanoMani or elsewhere — request data sheets that include these measurements. Suppliers who provide complete characterisation data stand behind their materials; those who do not may not know what they are actually selling.