Particle sizing? Try using nanosight!

Article image: 

Following on from my last blog on image use, I’d like to get back to exploring the reason we’re all here, microscopy!

Some of you no doubt will have received requests by potential users to use your EM kit as a particle sizing tool for the analysis of sub-micron particles. In my opinion whilst this can be done, it is time consuming and can often leave the operator, if not the sample owner, wondering how representative the sampling was. Specific concerns are the actual number of particles counted, fears over potential aggregation and other drying effects such as non-uniform sample dispersion over the grid (TEM) or stub (SEM). For these reasons quantitative measurement is typically better performed using light scattering techniques such as dynamic or static light scattering or techniques such as disc centrifugation. This blog appears on Microscopy and Analysis so I’m not going to cover light scattering techniques, but there is one technique that I think does seem to fit our brief.

About four years ago, I moved to the Nottingham Nanotechnology and Nanoscience Centre in Nottingham, where my main role was to operate a focused ion beam scanning electron microscope (FIB-SEM). However, in a previous life I was a polymer chemist and naturally did some sizing of polymer solutions and vesicle (liposome) suspensions. I was therefore the natural choice to oversee a recently acquired particle sizing instrument, a model LM 14 made by NanoSight, now part of Malvern Instruments.  

This system combines light scattering with a microscope and captures video, before analysis to give a particle size distribution. So, how does it actually work? For a full detailed version, head to the Malvern website  or click here for a great little video that explains it, but I'll give you the short version that seems to work for most people.

Firstly, the particles to be analysed must be in a solvent, which can be aqueous (most commonly), but could be ethanol, propanol, acetone, chloroform or hexane (I've used them all) or something else. Regardless of your choice, one thing you must know is the viscosity of the solvent at the temperature you measure the solution at. There are a range of models available with either manual or automatic temperature monitoring and/or control, however, the temperature and viscosity are crucial to accurate particles sizing and must be known!

Schematic of the NanoSight principle. Laser is scattered and the diffracted light is captured by a camera perpendicular to the fluid plane. Image courtesy of NanoSight

Once the solution containing the particles is injected into the viewing window, a laser is activated and the particles act as point scatterers. Next use either the oculars on the microscope or the attached camera, which feeds to the computer to adjust the stage to bring the scattering particles into focus. It is important to stress that we do not see the particles themselves, rather they appear as bright disks due to the diffracted light. In this way the LM 14 can visualise and record the scattered light from particles as small as 10 nm in the case of silver or gold, or around 30 nm for less strongly scattering particles. The instrument typically works well in the 20-2000 nm range. 

Now that things are focused, the camera and computer capture a video of the particles under Brownian motion, which can range in duration from 10 to 215 seconds. The video is then analysed once a few processing parameters have been set and the software, the Nanoparticle Tracking Analysis (NTA) tracks individual particles within the capture screen over the lifetime of the video or as long as they are visible, which may occur when moving off the screen or out of focus.

By using both the known values of temperature and viscosity of the solvent and by tracking the displacement in X and Y frame by frame, the hydrodynamic radius or volume can be calculated - in essence at a given temperature and viscosity smaller particles move further than larger ones. We can see how both of these values are used in the Stokes-Einstein equation along with the Boltzman constant:

The Stokes-Einstein equation, which links viscosity, temperature and displacement of each particle

Once the NTA software has analysed and calculated the Dt component (displacement) the Dh (size) can be calculated for each particle and a distribution of the sample displayed.

So, am I trying to say that my system is better for particles sizing than TEM or SEM? Well, that obviously depends on what you want to know. There are a few points to make: it should be clear that this is one of the techniques that is better suited to sizing particles in solution, since it eliminates fears over drying effects without the need to go to cryo-TEM or cryo-SEM, which can be time consuming (but do provide nice images). Compared with EM there can also be no doubt that a greater number of particles can be sampled in a shorter time (typically as 30–60 seconds!), which is typically enough to track in excess of 200 particles and conveniently the NTA software provides a graph of the data as size vs concentration, surface area or volume. This data is also available as both raw and processed data for offline analysis. 

One aspect that I haven’t touched on or used myself is the potential to use fluorescent particles with the LM 14. For a long time fluorescent tagging or expression of fluorescent proteins has enabled microscopy of particles, proteins or nanostructures. This doesn’t have to be limited to static samples and the system takes advantage of this. More recently quantum dots have been introduced to the market, which have significant resistance to bleaching and can be tailored to the application. Regardless of your choice the range of laser modules (red, green, blue) available and the addition of a filter enable tracking of these particles/proteins etc over non-fluorescent ones. This is a significant advantage over EM, where the fluorescent dyes are typically not an advantage, but are increasingly use by biology and medical researchers.

Particle size distribution profiles (yellow graph) of a mixture of 100 nm fluorescent and 400 nm non-fluorescent polystyrene particles analyzed under a) scatter mode and b) fluorescent (optically filtered) mode. Courtesy of Nanosight

An often overlooked aspect of the system is that it reports not only sample size, distribution and concentration, but also the intensity of scattered light. Whilst this might not seem terribly useful, there are occasions where it can be. For example imagine we have two similar sized particles of polystyrene and gold. The scattering intensity of gold would be higher as shown in the image, which would allow us to differentiate them based on this parameter. 

Within similar sized particles, higher scattering intensity can help differentiation of the species. Used with permission from Nanosight 

If we contrast this with EM, our options would be backscatter contrast in SEM to show the difference more clearly than secondary imaging (SEI) or high angle annular dark field (HAADF) in TEM to be more obvious than bright field imaging. Again the time taken to do this on either microscope would be greater than the Nanosight LM 14 and the previously mentioned worries over sampling still stand.

There are application and other aspects of the system that I haven’t touched on here without making this blog even longer. So far NanoSight's systems have been a contributing factor in excess of 1000 published papers in all manner of fields or research, some of which are reviewed hereHowever, next time you someone asks you about particle sizing, you might consider referring your client to someone with a NanoSight instrument and save the EM time for something else. Do you agree? Have you tried these instruments? I’d be keen to hear your thoughts.  

Website developed by S8080 Digital Media