How does Microfluidic Diffusional Sizing (MDS) compare to Dynamic Light Scattering (DLS) for protein size tests?

How does MDS compare to DLS for size tests

Measuring the size of proteins provides insights into folding and conformations, aggregation and oligomerization.

Dynamic Light Scattering (DLS) offers a fast and relatively simple method to determine size, but does have some technical drawbacks in terms of large particles in the sample biasing results.

Microfluidic diffusional sizing, as employed on the Fluidity One instrument, avoids the large particle bias, allows size measurement down to lower concentration levels, and is just as fast and easy to use.

What size range does each method cover?

DLS is able to assess protein size from 0.3 nm to 10,000 nm (10 µm), but this is sample dependent.

MDS can size proteins between 0.3 nm and 20 nm.

Both methods report protein size as hydrodynamic radius.


What sample quantity is required?

Sample requirements for DLS is instrument and cuvette dependent, with the smallest volume cuvette being 12 µL.

Comparison of the two methods using BSA indicates that samples must be ~41 µg/mL or more for DLS to offer accurate results, whereas MDS can produce results down to 13.8 µg/mL.

The detection chemistry used in Fluidity One will be saturated at concentrations higher than 500 µg/mL. DLS is capable of sizing proteins at concentrations as high as 40% w/v.


How does the workflow compare on each method?

In DLS sizing the sample is filled into a cuvette and placed in the instrument. Parameters are set in the linked software, and results are shown in 3 – 5 minutes.

For MDS tests 5 µL of sample is pipetted onto a disposable microfluidic chip, which is loaded into the instrument. The estimated size (from pre-set small, medium and large options) is chosen on the built-in touchscreen, then results are given in ~8 minutes.


How does each method work?

DLS uses the relationship between the size and speed of particles. As the molecules move in Brownian Motion in a solution, their size affects their speed – larger particles move slower.  A laser is shone onto the sample, and the light is scattered by the particles. The scattering intensity at a specific angle changes over time with the particle movement, and this is detected and related back to size using the Stokes-Einstein relationship.


In MDS, the sample enters a microfluidic channel on the measurement chip, along with an analyte liquid. As the system is in the micro range, the two streams do not mix and instead run side by side in a steady state laminar flow. During this time, protein molecules will diffuse from one side to the other, at a rate proportional to their size. After a short time, the streams are split again, and the proteins present in each side are labelled and detected. The amount of diffusion that occurred is then related back to size, again by the Stokes-Einstein relationship.


MDS Chip Flow schematic

The flow of sample through a microfluidic chip during MDS analysis on a Fluidity One instrument

Are there any limitations on either method?

DLS is biased towards larger particles, as the scattering intensity is proportional to particle diameter to the 6th power. This means if there is a small number of aggregates in a solution, this can lead to overestimation of the size.

If a mixture of proteins is present in a DLS test, they must differ in size by a factor of 2 (50%) or more in order for the 2 species to be resolved separately. This can result in some cases where a compact dimer and an extended monomer appear the same.

There are also some minor difficulties in preparation for DLS tests; sample dilution must be handled carefully - too dilute and the signal-to-noise ratio is too low, and too concentrated and multiple scattering effects or inter-particle interactions interfere with results. Preparation is sensitive also, dust or other contamination in cleaned and re-used cuvettes can affect results - though disposable cuvettes are available to avoid this.

When testing by MDS on the Fluidity One, the sample must not include any primary amines (inc. tris buffers) as these are also labelled and detected during the final stages of the analysis.

How do the results compare?

We measured a variety of proteins spanning a wide range of sizes. The results obtained show that both techniques generate consistent results, see Figure 1. These values are obtained using a Fluidity One instrument for MDS results, and a Zetasizer Nano ZSP for DLS results. Results represent an average of 3 readings.

MDS vs DLS for various size proteins
Figure 1: Results of sizing a variety of proteins spanning 3 orders of magnitude in size with both MDS and DLS

We next sized BSA over a variety of concentrations, using the same instruments. This showed that MDS was able to provide accurate sizing down to 13.8 µg/mL, far lower than the 41.7 µg/mL of DLS - see Figure 2.

MDS v DLS with BSA at changing concentration
Figure 2: Comparison of MDS to DLS - sizing BSA by both methods at various concentrations and comparing the measured size to the reported literature radius.

For full experimental details and discussion on these tests, see our recent application note here.

Summary of DLS vs MDS for protein size tests 



DLS - Dynamic Light Scattering


(Based on the Zetasizer Nano ZSP – other brands and models available)

MDS - Fluidity One
Measurement method Fluctuations in light scattering Microfluidic Diffusional Sizing MDS
Sample volume required 40 µL 5 µL
Measurement range

0.3 – 10,000 nm (10 µm)


(sample dependent)

0.3 – 20 nm
Measurement time 3 – 5 minutes ~ 8 minutes
Type of size reported Hydrodynamic Radius Rh Hydrodynamic Radius Rh
Can protein charge be measured?

Yes – by zeta potential measurement


(750µL of sample and specialist cuvette required)

No (will be possible on future instruments)
Reagents required? Some cuvettes disposable to prevent cross-contamination Disposable chips
Separate PC required? Yes No

Bias towards large particles – the overall size of a sample can be overestimated.


If 2 or more species present, they must differ in size by 50% or more to be resolved separately.

No primary amines (inc. no Tris buffers)
Is high throughput testing available?

Yes – with additional joined unit


For a more in-depth comparison of the techniques, complete with experimental data, see our application note here.

If you have any questions about protein size tests, just ask us here.

  • Publications and resources

    View all
    • Blog04 February 2019

      Biophysics for Biologists

      Protein X is a key suspect in causing disease Y. So how do you study protein X?

      Here we look at biophysics from a biologist's point of view; asking why is biophysics important? What can biophysics contribute to biology? And what biophysics techniques should biologists know about, and use?