The cumulant analysis and the size distribution (or regularization) analysis are two distinct methods to fit the autocorrelation function (ACF) obtained from dynamic light scattering experiments. Both analysis types provide useful information on particle sizes and the homogeneity of a sample, but in different ways and different levels of detail.

The cumulant fit models the ACF using a single averaged diffusion coefficient to obtain a single averaged hydrodynamic radius rH with a polydispersity index (PDI), while the size distribution fit allows for a distribution of particles with discreet hydrodynamic radii when modeling the ACF. In the tabulated results, the cumulant model reports a single rH and PDI, while the size distribution model reports the hydrodynamic radii and PDI for up to three discreet peaks (the model can fit more than three peaks, but only the three largest will be summarized in the results table). 

When analyzing DLS data, first analyze the results from the cumulant fit. If the rH is in the expected range and the PDI is sufficiently small, indicating a monodisperse sample, it is recommended to use these values for reporting particle sizes. A cumulant rH smaller or larger than expected together with a large PDI likely indicates a polydisperse sample. In this case, it is inappropriate to use the cumulant fit results. Instead, switch to the results from the size distribution fit. Only for perfectly monodisperse samples, the cumulant model and the size distribution model are expected to yield similar results.

Monomodal and multimodal distributions

Monomodal distributions are samples that have particles of only a single size or a very narrow size range, while multimodal distributions are seen in samples with multiple discreet particle sizes in solution. Monomodal distributions feature a single peak and are best analyzed and understood with the cumulant fit, while multimodal distributions feature multiple peaks and are best analyzed with the size distribution (regularization) fit. In more practical terms, this means that if you want to detect aggregates or vice versa fragmentation and if you are interested in the number and the relative composition of populations in your sample, the size distribution results will be more helpful.

In addition, Prometheus Panta software helps with the decision of which fitting model to use through automatic quality checks: If the warning for the cumulants analysis appears, the two fitting results should be inspected carefully and you may consider relying on the results of the size distribution fit. It can also be helpful to observe the fit residuals displayed below the ACF in the acquisition details tab to judge which of the two fitting models better represents the data.

Be aware that scattering data is only available if data was generated with measurement types supporting DLS and with Dynamic Light Scattering set to on.