In DLS method, the time dependence of scattered light intensity is used to obtain autocorrelation function (usually characterized by about 100 points). Then this function is fitted, and the inflection point of the fit corresponds to the characteristic time of the process in the sample. If we assume that this process is the Brownian motion of particles, then we can calculate characteristic size of these particles (regarded as hydrodynamic size; assuming spherical particles, and taking into account viscosity and refractive index of the medium).
Is it possible to achieve the estimate of this particle size with accuracy better than 1 nm for particles, say, of 10 nm? I suspect that it should be inherently limited by accuracy of the fit for the autocorrelation function.
The correlation plot is an exponential decay and thus is very low resolution (only 2 parameters needed to define/model it), so the '100 points' in your commentary doesn't imply more accuracy or precision - a better fit doesn't necessarily result. We also note that size standards in the < 100 nm range tend to be characterized by electron microscopy, so there is not a like-for-like (intensity and number) comparison. We also note that the approximate size of a hydrogen atom is 74 pm (or about 0.07 nm), so the best we would hope to achieve would be quoting to +/- 0.1 nm. However, DLS measurements of materials like sucrose (sub-nm) by DLS are supported by other molecular data. See the attached.
We also have to be aware that we're relating a diffusion coefficient to size (via Stokes-Einstein) and thus any covering stabilizers and surfactants are factored into the hydrodynamic size. And 'size' is an ambiguous term for irregular particles.....SAXS or electron microscopy shows the electron-dense region of the particle and thus will not consider or 'see' the adsorbed layers. I think 10 +/- 0.1 nm is probably the best we could hope to achieve (with all the notes and caveats above). However, we must be positive about the DLS ensemble technique - how else can we get information from billions of particles simultaneously with great sensitivity in the intensity distribution to the presence of tiny amounts of aggregate?
The certificate for the RM8011 standard from NIST is enlightening in this respect.
Alan F Rawle , thank you for the information, and for very interesting paper. Probably, we can employ several well-defined water-soluble substances as calibration references.
Nature provides some nice materials to test such as BSA, hemoglobin, as well as sucrose.
- Badges
- Science topic
- Similar topics
- Agriculture
- Horticulture
- Seeds