How to Choose a Particle Size Analysis Method

Real powders and suspensions are not spheres. Every technique therefore reports an irregular particle as an "equivalent spherical diameter" defined by the physical quantity it actually measures. Laser diffraction reports the diameter of the sphere that would produce the same scattering pattern, microscopy reports a projected-area equivalent, and sieving reports the smallest aperture the particle can pass through. Different numbers for the same sample are not an error; they are a consequence of the method.

The second point is weighting. Laser diffraction produces a volume-weighted distribution, so a handful of large particles pulls the D50 upwards noticeably. Microscopy and dynamic light scattering produce number- or intensity-weighted results that make fine particles far more visible. A technically sound report should therefore state not only the D50 but also D10, D90, the span of the distribution and which weighting basis was used.

Before choosing a method, we recommend answering three questions. What size range do you expect? Will particle shape or agglomeration influence the result? Will the data be compared against a specification, or used to rank batches against each other? In most cases the answers point directly to the right technique, and they also tell the laboratory which dispersion conditions and reporting format will make the result useful to you.

Laser diffraction (ISO 13320) covers a range from roughly ten nanometres to several millimetres in a single measurement and delivers a statistically robust, volume-weighted distribution within minutes. Wet dispersion handles suspensions, emulsions and powders that can be dispersed in water or an organic solvent; dry dispersion handles free-flowing powders. Its high repeatability makes it the natural choice for quality control and specification compliance.

The critical input is the optical model. Mie theory requires the refractive index of both particle and medium together with the absorption coefficient of the particle; for sub-micrometre material an incorrect index can shift the fine end of the distribution substantially. The Fraunhofer approximation is acceptable only for large, opaque particles. A sound report therefore states the optical model and the dispersion conditions used, including solvent, sonication time and stirrer speed.

Laser diffraction gives no information about particle shape and treats a tightly bound agglomerate as one large particle. An unexpectedly high D90 is more often a sign of incomplete dispersion than of genuinely coarse material. Cross-checking against SEM images is the most reliable way to resolve that ambiguity, and repeating the measurement after additional sonication quickly shows whether the coarse tail is stable or collapses.

Scanning electron microscopy (SEM) images particles directly. Beyond size, it reveals shape, surface texture, the degree of agglomeration and the difference between primary particles and secondary clusters at a glance. For platelet, needle-like or fibrous materials the equivalent diameter from laser diffraction has little physical meaning, and image-based length and aspect-ratio measurements are far more informative.

The price of a size distribution from SEM is statistics. A number-weighted distribution that follows ISO 9276 principles requires hundreds of particles, and thousands for broad distributions, while sample preparation (dispersion, coating) influences the result. SEM is therefore rarely used as a stand-alone routine quality-control method; its real value is as a reference technique that validates and explains laser diffraction data.

At the nanometre scale SEM is often the only accessible direct method. A primary particle size below one hundred nanometres should never be confused with the hydrodynamic diameter obtained by light scattering; the difference between the two values is itself informative, since it reflects the stabiliser layer on the surface and the state of agglomeration in the dispersion.

Dynamic light scattering (DLS, ISO 22412) derives the hydrodynamic diameter from Brownian motion in dilute colloidal suspensions and performs best from a few nanometres to roughly one micrometre. Because it is intensity-weighted, a few large particles or traces of dust can dominate the result, and a single mean value is misleading for samples with a high polydispersity index. DLS is the typical choice for liposomes, polymeric nanoparticles and protein solutions.

Sieve analysis, using test sieves to the ISO 3310 series, occupies the coarse end of the range. For free-flowing granules and powders above approximately 45 micrometres it is economical, mass-weighted and easy to interpret, and the result maps directly onto industrial classification processes. Moist, electrostatic or platelet-shaped materials, however, cause sieve blinding and orientation effects that introduce serious errors.

If the expected range lies between 0.1 and 1000 micrometres, the sample can be dispersed and the purpose is specification control or batch comparison, laser diffraction is the first choice. Add SEM when the particles are clearly non-spherical, agglomeration is suspected or the primary particle size is below one hundred nanometres. Use DLS for dilute colloidal suspensions below one micrometre and sieving for granules above 45 micrometres. In most development projects two complementary methods used together yield far more insight than any single one.

Whichever method is chosen, the report should document the sampling and sample-splitting procedure, dispersion conditions, optical model, number of repeat measurements and the weighting basis of the D10, D50 and D90 values. Without these details the results of two laboratories cannot be compared. At NAKTEM we evaluate the choice of method together with your sample and, where it adds value, present laser diffraction and SEM results side by side in one report.