Measuring Particle-Size Distributions

Introduction

A nanoparticle batch almost never contains one exact diameter. “The size is 10 nm” hides the width, shape and weighting of a distribution. Counting individual cores by microscopy tends to produce a number-weighted view, while optical scattering can give disproportionate influence to larger objects. A material's total surface or mass may also be dominated by different size classes. Correct interpretation starts by asking which physical quantity each measurement weights.

Core explanation

A number-weighted histogram counts every detected particle once. If 90 particles have 5 nm diameter and 10 have 20 nm diameter, the number fractions are 90% and 10%. The number mean is (90 × 5 + 10 × 20)/100 = 6.5 nm. That value describes a typical counted object, but it does not describe where most mass is. For spheres of one density, surface area scales as diameter squared and volume or mass scales as diameter cubed.

Using the same mixture, total relative area from the 5 nm group is 90 × 5² = 2,250 arbitrary units, while the 20 nm group contributes 10 × 20² = 4,000. The minority large particles supply more area than the majority small particles. Relative volume contributions are 90 × 5³ = 11,250 and 10 × 20³ = 80,000, so the large group dominates mass even more strongly. A number-only statement can therefore conceal the size class responsible for exposure, dosage or material consumption.

Dynamic light scattering, DLS , estimates diffusion behavior from fluctuations in scattered light intensity in a liquid. It commonly reports an intensity-weighted result and a hydrodynamic diameter. For sufficiently small ideal particles under Rayleigh-like conditions, scattering intensity rises very steeply with size, roughly with the sixth power of diameter, so a few large aggregates can dominate the signal. The exact weighting depends on optical properties and regime. A large DLS mode is an important warning about aggregation, but it should not be translated directly into a number fraction without a justified model.

The hydrodynamic diameter includes the effect of solvation and surface coatings as the object diffuses. It need not equal the dry crystalline core diameter seen by TEM. A polymer-coated 10 nm core may move like a larger object in water; an aggregate of several cores may move like a much larger object. DLS conversion from intensity to volume or number weighting requires refractive-index assumptions and inversion of a noisy distribution. A presented “number DLS” plot can look reassuring while obscuring the raw intensity contribution of aggregates.

Microscopy is not assumption-free. A TEM analyst may count only particles that are visibly separated, omitting aggregates. Small particles can be hard to detect against the grid; large ones can be split or merged by segmentation. Report the image threshold, sampled fields, particle count and treatment of touching objects. AFM can provide height-based dimensions but lateral widths are tip broadened. Small-angle scattering samples large ensembles and can provide core distribution estimates, yet fitting requires assumptions about shape, contrast and interactions.

Distribution width matters as much as mean. Two batches can share a 10 nm number mean, one narrowly centered near 10 nm and another mixing very small and large particles. Their optical absorption, sedimentation and catalytic behavior may differ. Report median, spread or percentile values and show the histogram when feasible. For strongly skewed distributions, a single mean may represent no common particle at all.

The weighting chosen should match the application. Number weighting helps answer “what proportion of objects are in each class?” Area weighting helps estimate total reactive interface if all surfaces are accessible. Mass or volume weighting helps determine where most material resides. None of these automatically predicts toxicity or catalytic rate because chemistry and accessibility also matter. But choosing the correct basis prevents basic denominator errors.

Step-by-step reasoning

Define whether the desired measurand is core diameter, hydrodynamic diameter, aggregate size or functional surface. Record the raw method and distribution basis. Inspect full histograms for secondary modes. Cross-check a count-based method such as microscopy against a liquid method such as DLS or SAXS, accounting for sample preparation. Convert weighting only when shape, density and optical assumptions are justified. Report uncertainty and the number of independent batches.

Visual explanation

Draw 90 small circles and 10 large circles. In a number bar chart the small group towers over the large; in an area chart the large group becomes prominent; in a mass chart it dominates. Beside this show one coated core with a TEM core-diameter line and a larger DLS hydrodynamic-diameter circle including its solvent-bound shell.

Real-world analogy

A town may have many small houses and a few warehouses. Counting buildings says the houses dominate; counting floor area or stored goods may say warehouses dominate. Particle distributions behave similarly under number, surface and mass weighting. The analogy reminds us to state the denominator before comparing “average size.”

Real-world example

A batch's TEM count histogram is narrowly centered at 8 nm, yet DLS shows a broad mode above 100 nm. This can happen if a small number of aggregates strongly influence scattering or if drying on TEM grids favored isolated cores. The appropriate response is to check the dispersion in its actual medium, perhaps with dilution and imaging controls, rather than automatically declaring either instrument wrong.

Why?

Size distribution controls how a nanomaterial behaves in light, flow, reaction and exposure. Comparing unlike weighting bases can create false contradictions between instruments or hide a large-particle tail. Transparent reporting lets others calculate which part of a batch governs a specific property.

Common misconception

“A DLS mean is the same as a TEM mean” is false: the techniques often measure hydrodynamic versus dry core size and weight populations differently. Also, a monomodal DLS result does not prove perfect monodispersity; inversion resolution and signal weighting can hide minor populations.

Worked example

Suppose a batch contains 99 spheres of diameter 5 nm and one sphere of diameter 50 nm. The number mean is (99 × 5 + 50)/100 = 5.45 nm. Relative volume of the small group is 99 × 5³ = 12,375, while the single large sphere contributes 50³ = 125,000. That one object accounts for about 91% of the total volume in this simplified set. The worked calculation shows why rare large particles matter to mass and scattering.

Quick check

1. Why can a tiny number of aggregates dominate a DLS intensity signal? Answer: Large objects scatter much more strongly than small ones, especially in the small-particle optical regime.

Exam focus

State whether a reported distribution is number, intensity, area, volume or mass weighted. Use diameter squared and cubed scaling for similar spheres. Distinguish dry core size from hydrodynamic size. Do not compare means across methods without checking sample state and weighting.

Advanced insight

For polydisperse systems, different mathematical means emerge from different moments of the size distribution. Transforming an intensity-weighted DLS inversion into number weighting can amplify uncertainty in small-particle bins because it divides by a steep size-dependent scattering response. Multimodal populations challenge inversion resolution, so orthogonal methods and reference materials improve confidence.

Summary

Particle size is a distribution with a defined measurement basis. Count, surface, mass and scattering signals emphasize different size classes, while hydrodynamic and dry core diameters describe different physical objects. Report full distributions, weighting, preparation and uncertainty. Use the basis appropriate to the question rather than one unqualified “average.”

Practice questions

1. A sphere doubles in diameter. How do its surface area and volume change, assuming the same shape? Answer: Surface area rises by a factor of four and volume by a factor of eight. 2. Why might DLS report a larger diameter than TEM for a stable polymer-coated particle? Answer: DLS reflects diffusion of the coated and solvated object, while TEM may show mainly the dry inorganic core. 3. What is the number mean for eight 5 nm particles and two 10 nm particles? Answer: (8 × 5 + 2 × 10)/10 = 6 nm. 4. Which weighting is most directly relevant to how much material mass each size class contains? Answer: Volume or mass weighting, with appropriate density information.

Sources: NIST study on DLS distribution weighting; NIST DLS nanoparticle measurement protocol; Primary light-scattering study of size-dependent signal.