Dynamic light scattering data interpretation using the Zetasizer Advance series

1. Introduction

This technical note focuses on interpreting data and results obtained from size measurements using the ZS Xplorer software for the Zetasizer Advance series of instruments. Readers will learn how to identify good and poor-quality data and methods to improve data quality will be discussed.

The quality of data from size measurements is essential for ensuring result reproducibility, since higher-quality data leads to more consistent measurements. It is equally important for generating representative and meaningful results that can be effectively used in a wider study. To facilitate this, we highlight various parameters and charts that can be used to assess the quality of DLS (Dynamic Light Scattering) results.

2. Decoding the correlogram

When evaluating DLS results, the optimal starting point is often the correlogram. Correlation is a statistical method that analyzes fluctuations in the intensity of light scattered by particles over time. It is used to determine how quickly the intensity of the signal fluctuates, which is directly related to particle size. Therefore, in a DLS measurement, the correlogram contains a lot of information about the sample, making it a valuable resource for data interpretation and identifying quality issues. An example of a typical good quality correlogram is presented in Figure 1:

As seen in Figure 1, the correlogram can provide a wealth of information about the sample. For instance, the time that the decay starts indicates the mean size. The smaller the particles, the more rapidly the correlation decays. Additionally, the gradient of the decay indicates the polydispersity of the sample. Samples with low polydispersity exhibit steep, sharp gradients, whereas more polydisperse samples display a more gradual, shallow gradient.

Data quality information can be assessed by analyzing the correlogram. For example, the y- intercept is representative of the signal to noise ratio of the measurement and its value will vary according to the sample and instrument optical configuration. The closer it is to 1, the better the quality of the measurement, however any values between 0.1 and 1 are appropriate. A lower-than-expected value may indicate too high or too low a sample concentration, absorption, or fluorescence of the sample. Too high a sample concentration may lead to multiple scattering effects that reduce the y-intercept value (as shown by correlogram in Figure 1A). The back scatter configuration of the Zetasizer Pro and Ultra can allow for the path length of the laser beam to be minimized (Non-Invasive Back Scatter, NIBS¹), which, in turn, reduces the occurrence of multiple scattering.

The baseline of the correlation function can also be used as a quality indicator of a DLS measurement. Samples suitable for DLS measurements should achieve a flat baseline as shown in the orange correlogram in Figure 1B. However, if a sample is not suitable, for example when number fluctuations or particle sedimentation are present, the baseline can be elevated and appear noisy even at long delay times as shown by the blue correlogram in Figure 1B.

3. Number Fluctuations

When larger particles are present in the sample, for example, dust or aggregates, they may intermittently enter and exit the measurement volume as they diffuse under Brownian motion, during the measurement, causing the intensity of the scattered light to fluctuate. This phenomenon is called number fluctuations.

Number fluctuations can decrease the accuracy and repeatability of the results. The primary consequence of number fluctuations is typically a raised baseline, as illustrated in Figure 2A. In extreme cases, when the number fluctuations are large, the intercept will vary and may even exceed 1 making the result unreliable as it is depicted in Figure 2B.

To recognize number fluctuations, in addition to looking at the correlogram, one can also look at the In Range (%) parameter.

The In Range (%) parameter provides an overall indicator of data quality. It is derived by comparing the measured baseline (the experimental baseline) with the theoretical (predicted). It is scaled from 0 to 100% and the more similar the measured baseline is with the theoretical, the higher the In Range (%) parameter will be. Ideally, In Range (%) values should be as high as possible e.g., >90%. Values <90% are typical of samples which exhibit number fluctuations. Monitoring the intercept and In Range parameters, is a good way to evaluate the quality of a series of measurements. In Figure 3, a couple of examples are highlighted where the intercept exceeded 1 or was low, paired with low In Range (%) values. These parameters can be added to a parameter table in ZS Xplorer software, as discussed in section 5.1.

Number fluctuations caused by the presence of large particles such as dust or aggregates, can be reduced by filtering the sample through an appropriate pore size filter, centrifuging the sample, or allowing the large material to naturally sediment over an extended period prior to measurement.

4. Results repeatability

In addition to the quality of each individual result, the repeatability across multiple results should also be monitored to ensure there are no changes occurring in the sample during the course of the measurement. Several repeat measurements should be performed per sample, with a minimum of 3, preferably 5, recommended.

Correlation functions from repeat measurements ideally should closely overlap each other as shown in Figure 4A. This indicates that the sample is stable over the duration of the measurements. If the sample is changing with time, the correlograms will not overplot each other closely (Figure 4B) indicating that sedimentation, creaming, aggregation or particle dissolution may be occurring.

The Z-average diameter as well as the mean count rate across the repetitions should also be examined.The Z- average is the intensity weighted mean diameter of the particles in the sample. The mean count rate is the average count rate obtained during the measurement.

For a suitable DLS sample, meaning a relatively monodisperse sample within an appropriate size range, measured at a suitable concentration, the Z-average results should be within 2% of each other. However, polydisperse samples are expected to have more variable results. Variable and trending results are another indication that the sample is changing over the duration of the measurements and can be caused by aggregation, dissolution, sedimentation or creaming.

4.1 Aggregation and Dissolution/Sedimentation

To understand whether aggregation is occurring during a DLS measurement, the Z-average diameter and the Mean Count Rate can be checked, as shown in Figure 5.

Figure 5 is an example of a noticeable increase in both the Z-average diameter and mean count rates. As particles aggregate, this results in an increase in the amount of scattered light, which is reflected by increasing mean count rates.

Conversely, Figure 6 shows decreasing Mean Count Rates and Z-average diameters indicating that sedimentation, creaming or dissolution is occurring.

If instability is observed upon repetition, the sample may need to be prepared in a way which stabilizes the particles. For example, adjustment of the sample pH, or the addition of an additive such as a surfactant, could help with sample stabilization.

5. ZS Xplorer aiding with data interpretation

Throughout this technical note a few charts and parameters have been mentioned that can aid with data interpretation. All these parameters and charts plus more are available in ZS Xplorer software. These are all summarized below as well as how to add them to the workspaces in the ZS Xplorer.

5.1 Recommended Parameters

Table 1 shows a list of recommended parameters that are available in ZS Xplorer software which can be used to aid interpretation of the data and results obtained.

Parameters can be added to the Statistics or Parameter tables in the software. To do this, users can click on the icon positioned at the top right-hand corner of a table,   followed by the “Selects the displayed properties” icon.   From the “Available parameters” list (Figure 7), the parameters required can be selected and inserted into the table and reordered if required using the arrow buttons. 

Changes in the Size workspace are only effective while that workspace is open. Changes in the Custom workspace, however, are saved even after exiting the software.

5.2 Recommended Charts

Table 2 shows a list of recommended charts that are useful in aiding the interpretation of size measurements. Charts can be selected from the drop down lists available.

6. Data Quality Guidance

For size measurements, a trained neural network is employed to evaluate the data quality. It examines the correlogram, considers various factors such as shape, intercept, and measurement angle, to alert the user to the quality of the measurement. This assessment may be communicated through one of three icons presented in Table 3:

Table 3: The records view icons depending on the quality of the measurement.

Records view icon	Interpretation
	No data quality issues found.
	The data is usable, but advice should be followed when using this data. e.g. Multiple populations – the distribution analysis is a meaningful description of the results, and the Z-average cannot be easily interpreted
	The data is of poor quality and the remedial advice should be followed before using data from this sample

Different icon colors are assigned to the measurements depending on the quality of the data:

• A solid green icon featuring a white tick signifies that no data quality issues were detected.
• A white icon with a green outline indicates that the data is usable, but caution should be exercised, and any accompanying advice should be followed.
• A blue icon notes that the data quality is poor, and remedial advice should be taken before utilizing the data from the sample.
• There are also grey icons for situations where advice is unavailable, such as for average records.

As showcased in Table 3, in addition to highlighting issues to the user about the quality of their measurement, the Data Quality Guidance also provides advice on how the quality could be improved and how the results should be used. For example, the hollow icon in Table 3 notified the user that their sample was polydisperse and that the distribution analysis would be more robust in describing and reporting this result. The advice could also include a suggestion to try a Multiple Angle Dynamic Light Scattering⁶ (MADLS) to increase the resolution of the distribution.

7. Conclusion

This technical note provides a comprehensive guide on interpreting DLS data. It covers key aspects such as analyzing the correlogram and how to look for trends within repeat measurements. The note also summarizes key parameters and charts that can aid with data interpretation and assist users to gain insights into the characteristics of their samples.

8. Further reading

1. Dynamic Light Scattering – an introduction in 30 minutes
2. What is a Correlation Function (Correlogram)?
3. Dynamic Light Scattering – Common terms defined
4. Explaining the difference between Non-Invasive BackScatter (NIBS) and Backscatter – they are NOT the same.
5. Dynamic Light Scattering method development using the Zetasizer Advance Series
6. Improved component resolution with Multi-Angle DLS (MADLS)