The above answers are all correct. In practice the limit depends on the amount of dust and dirt in the sample. Remember, scattering power goes by the 6th power of particle radius (at small angles). Therefore it is rather difficult to measure e.g. a protein like b-lactoglobulin due to the presence of a very few aggregates or dust particles. So in theory no problem. But in practice not easy. Because you do not know whether dust etc was removed

It is not so simple, depends on the wavelenght of laser, power, sample (number of particles, size, refractive index..).

Most of equipments of DLS claim to be able to get values of 1 nm. In my oppinion this is the very very low limit

As DLS is probing relaxation times of concentration fluctuations at a given inverse length scale (scattering vector q=4pi/lambda*sin(theta/2) in m^-1), these relaxations have to take place within the time window of the correlator. Assuming a monodisperse solution of particles, calling R the characteristic radius of a particle and C its concentration, in the limit where q*R

Resolution of DLS instruments is limited by wavelength of light (laser). It may not be good idea to probe length scale about 1-5nm using DLS. I think small angle x-ray scattering would be appropriate technique to probe such a length scale.

DLS is often called QELS (quasi elastic ligth scattering) or PCS (photon correlation spectroscopy). It monitors fluctuations of the scattered intensity (or electric field) by the solution. These fluctuations being related to concentration fluctuations in the solution . Using "dynamic" X-ray scattering (XPCS) is only possible on large facilities (like ESRF in Europe) and is not accessible in a regular lab. Dynamic scattering techniques give information about dynamics of the solution and under certain conditions give access to the hydrodynamic radius of particles. The wave length (lambda) is important but the angle of measurement (theta) is also very important. The combination of both determines the accessible length scale (q^-1) and then the relaxation time measured for a given diffusion coefficient D (see equations in my previous message). The main limit is from my point of view the sensitivity of the correlator. What is the minimum relaxation time that can be measured with a given apparatus ? This is the main limit. Considering the minimum tau, you have a combination of D and q (using previous equations) that could correspond to it and you can derived the minimum size measurable. Any way, you have to be sure that the measurement of tau is made in the limit q.R

Depends on several factors as the wavelenght of laser, refractive index, power, etc

Our equipment is to be able to get values from 5 nm.

The above answers are all correct. In practice the limit depends on the amount of dust and dirt in the sample. Remember, scattering power goes by the 6th power of particle radius (at small angles). Therefore it is rather difficult to measure e.g. a protein like b-lactoglobulin due to the presence of a very few aggregates or dust particles. So in theory no problem. But in practice not easy. Because you do not know whether dust etc was removed

There is no problem going down to very small particles, say, a few nm, with QELSS. For example, serum albumin, radius 5nm is easy. The main r\problem is the amount of scattered light. If you have a gross index mismatch, e.g., gold sol, life is easy. If the object is index matched to the solvent (see e.g. papers by Pusey on colloidal spheres, or papers by Lodge and others on polymers index matched to solvent, references in my book Phenomenology of Polymer Solution Dynamics, the colloids and tracer diffusion chapters) you may not be able to see it at all. One challenge is sample cleaning, aided by staying at larger angles. A larger laser is better. Shorter wavelength helps, but less than you might think, because signal to noise is photons per correlation time.

Hazard: QELSS measures the mutual not the tracer diffusion coefficient. See by papers in J Chem Phys, 1974. If you run up the concentration, D does not convert to particle radius.

For rigid aprticles qR

For nanoparticles with a liquid lipidic core surrounded by pegylated surfactants the criterium qR

For particles that are not very tiny, there is also an issue that the particles, even if completely monodisperse, may scatter different amounts of light toward the detector depending on their orientation relative to the source, detector, and light polarization. They may also have different diffusion coefficients, depending on which way they are oriented relative to the scattering vector (this issue does not arise for spheres). The result is that the apparent D depends on angle, because in some directions the particle scatters a lot of light when its slow axis is aligned parallel to the scattering vector, and in other directions the particle scatters a lot of light when the fast axis is aligned parallel to the scattering vector. This issue is discussed in Berne and Pecora's book, at least for a few simple shapes, though the phrasing 'rotational' coupling might be alternatively understood if you called it 'orientation' coupling.

I am very interested to read that there are particles that are soft enough to give the described effect. One might also see this with liposomes.

To follow the previous answer. To avoid aggregates and dust I would recommend to do the measurement in line with size exclusion chromatography that should remove all the potential dust and aggregates.

I think in the visible rang is from 3nm to 3micrometer

I am working with proteins. If well purified, hydrodynamic radii of 1-4 nm are accurately measured. Osteopontin (monomer) runs at 2.8-3.1 nm Rh. I am operating a DynaPro-E series instrument (ProteinSolutions/Wyatt).

I recommend George's answer

Best Regards

DLS can give quite reliable data, even down to less than a nanometer [ Article Measuring sub nanometre sizes using dynamic light scattering