No matter what your product, from injectables to tablets to freeze-dried materials, the size of the particles is critical. From “there shouldn’t be any in clear injectables” through micronized low solubility APIs through the effects of excipient particle sizes on solid dosage forms, knowing (by actual measuring) the size (and/or presence) of particles is a critical piece of data.

Being able to qualify the particles is as critical, too, however, this column will stick to the measurement of the size, not any physical attributes. There are a number of techniques for estimating/measuring the mean particle size of a solid material. Some are old and some newer, including: Nested sieves (e.g., USP method); Coulter counters; microscopy (photo-microscopy); LASER-light scattering (in a suspension); diffuse reflection (e.g., NIRS); and spatially resolved dynamic light scattering (SR-DLS).

The “tried and true” method of sieving a portion of the powder has been around for more than 50 years. The bulk material is poured onto the largest-holed sieve and the stack of steadily decreasing hole sizes is shaken/tapped for a set period of time. The amount (weight) of material retained on each sieve is measured and the weight-averaged particle size mean is calculated. One potential source of error is that the amount retained on the bottom pan is designated “less than the smallest sieve size.” That, alone, skews the actual value of the mean size.

Another flaw is a one sieve measurement—as per several USP monographs—for a raw material. The pass/fail criterion might read, “no more than 1% is retained on a 100-mesh screen.” When I was early in the process of generating my library of raw materials for NIR qualification, I examined a sample of Na Pentobarbital that had just passed QC. My newly minted NIR algorithm failed the sample. Upon examination, it was found to be micronized, not granular, The criterion for particle size? Less than 1% retained on a 100-mesh screen—easy to pass when micronized.

The referee method for the quandary was photo microscopy. The drawbacks to this technique are obvious: a) it is slow and b) it only shows a tiny portion of the sample. Most sampling techniques are suspect, anyway, but taking less than one gram from a multi-kilo sample hardly gives a credible value for mean particle size. It will remain a nice referee technique, but unlikely to replace other QC methods anytime soon.

These counters, where the slurry of particles flows through a “gate” are estimated as to size and number. This was popular in the 1970s but had several flaws, not the least of which was that an elongated crystal could be counted as several particles as it passed through the portal. These were supplanted by a LASER-light scattering device (Malverne) where the slurry of particles is stirred, and the diffusion of light is captured on concentric rings (light scatters more as the size decreases).

LASER-light scattering (LLS) instruments really came into their own in the 1980s. They were simple, but effective. The sample is suspended in a saturated solution of that same material to avoid dissolution and errors in measurement, stirred gently to assure distribution and avoid precipitation, and measured. While this was far better than earlier methods, there were still several caveats to consider, based, as always, on physics.

Anyone who has performed gravimetric analyses knows that you allow the precipitate and supernatant to sit overnight before filtering and measuring the material. This “digestion” is, in reality, the kinetics of dissolution and re-precipitation. This dynamic is continuous, but, based on surface area, the smaller particles disappear while the larger one gather mass. This process occurs in the LLS instrument. Timing is critical when measurements are being made—the longer you stir and measure, the more the mean shifts. In addition, since there is no primary reference for these data, scrupulous attention to technique will allow precision on sample-to-sample measurements.

The other mechanism involved is mechanical; in the stirred sample, larger particles will be beaten to smaller and smaller particles. This skews the measurement and the major reason for non-reproducible results.

The only alternative to measure dry powders by sieve analysis would be to use diffusely scattered light. My experience with NIR allowed me to graphically estimate the mean particle size of a powdered sample. The method works assuming a) a well-sampled bulk material and b) an equation, previously generated, is in place. The calibration, in this case, was made by sieving a bilk sample and measuring the cuts via an LLS device and creating a calibration graph (Absorbance at a specific wavelength versus the reciprocal of the mean particle size; A v. 1/µ). The Absorbance of the QC sample is then measured, and the particle size calculated. While there are, again, no reference standards, although there are sources for “known” polymer beads that could be used for reference, there are two advantages: a) the sample is stable and b) multiple readings (shake sample and re-read) may be made and multiple samples taken, since the method is so fast and non-destructive. At the recent IFPAC conference (N. Bethesda, MD), I came across another method for testing that was impressive. Spatially Resolved Dynamic Light Scattering (SR-DLS), recently developed as the NanoFlowSizer, employs SR-DLS, based on Low-Coherence Interferometry (LCI). It provides new possibilities for non-invasive, real-time, and continuous inline measurement of size characteristics in flowing and quiescent nano-dispersions. In Conventional Diffuse Light Scattering (DLS) measurements need to be performed under static conditions ensuring that particle movements are solely caused by Brownian motion and not influenced by other factors like liquid flow.

Additionally, conventional DLS cannot be applied to relative turbid suspension without dilution, while these are often encountered in industrial or process environments. Since nanosuspensions are in motion during processing and vary in turbidity levels conventional DLS is unsuitable for process analytical (PAT) applications.

Low Coherence Interferometry (SR-DLS) allows particle size characterization in process flows since it can measure highly turbid suspensions without dilution. The technology is based on low coherence interferometry providing light scattering information as function of optical pathlength (pathlength or depth in the sample). The sample is illuminated by low coherence light from a broadband source instead of a laser, and back scattered light interferes with light split from the source with a specific optical pathlength. The interferometer part of the technology allows to resolve backscattered light for specific path lengths in the sample simultaneously. The depth resolved light scattering data holds information on particle movement caused by both Brownian motion as well as flow rate and pattern. The contribution due to Brownian motion is extracted by smart algorithms and used for calculation of the particle size characteristics, while the flow rate information is obtained instantaneously for every measurement as well.

Nanoparticles are often seen in vaccines, where the size of the materials are critical in having the proper bioavailability and dosage levels. Since the process cannot be stopped for each measurement, SR-DLS allows real-time measurement and, if needed, changes to the process. The particle size data may be plotted over time to generate a “process signature” for each lot. The effectiveness of each lot is determined, and the 3-D plot may be used as a template for later batches. 

---

Emil W. Ciurczak, DoraMaxx Consulting, has worked in the pharmaceutical industry since 1970 for companies that include Ciba-Geigy, Sandoz, Berlex, Merck, and Purdue Pharma, where he specialized in performing method development on most types of analytical equipment. Email: [email protected]