Lab to line particle sizing for granulation optimisation

Laser diffraction and spatial filter velocimetry offer an alternative to sieve analysis for granule sizing and granulation monitoring. Paul Kippax, Malvern Instruments and Tamal Mukherjee, Malvern-Aimil Instruments, describe how they can be used

Fluidised bed granulation processes enable the conversion of fine, poorly-flowing tableting blends into a granulated feed that performs well in the tablet press. Such processes are complex, with mixing, wetting, granule growth and drying all taking place within the same piece of equipment. As a result it can be difficult to control the properties of the exiting granules, which are influenced by a number of factors, including raw material quality, processing conditions and design features of the processing equipment. Granulation scale-up is widely recognised as being particularly challenging.

The Quality by Design (QbD) approach, now widely adopted across the industry and required by regulators for generic submissions, calls for a detailed understanding of the factors that influence the Critical Quality Attributes (CQAs) of a finished tablet. These CQAs include dose uniformity, hardness and stability – parameters that can only be measured post-tableting. In contrast, granule size, which has been shown to correlate directly with certain CQAs, enables measurement much earlier in the tablet production process1. Granule sizing can therefore be extremely valuable for scoping the design space of a granulation process and for achieving reliable scale-up.

This article looks at the benefits of replacing traditional sieve analysis with modern laser diffraction particle sizing and spatial filter velocimetry (SFV) techniques, to monitor and control granulation processes efficiently.

While sieving remains in widespread use it is a slow, labour-intensive analysis that offers only poor resolution and risks granule break-up

While sieving remains in widespread use it is a slow, labour-intensive analysis that offers only poor resolution and risks granule break-up. Its replacement with laser diffraction and SFV enables faster granule size measurement. Experimental data confirm the close agreement of results measured using all three of these techniques, underlining the practicality of applying laser diffraction and SFV in tandem, in place of sieving, for fast, reliable granulation development, monitoring and control.

An introduction to fluidised bed granulation: Efficient handling of the fine powders that frequently constitute tableting blends can be problematic. Materials in the sub-10 micron (µm) range, where active ingredients often lie, are especially challenging because of strong attractive interparticle forces that result in high cohesivity and poor flowability. Granulating a tableting blend eliminates the health and safety risk associated with very fine particles, improves flowability and enables the close control of properties such as bulk density, compressibility and solubility, which can directly influence finished tablet CQAs. Furthermore, granulation can help prevent the segregation of very low quantities of active ingredient, thereby safeguarding content uniformity.

In fluidised bed granulation, a binder solution is sprayed onto particles that are suspended in an upward-flowing heated air stream. Larger sized agglomerates form from the fine powder as a result of liquid bridging. The process enjoys widespread industrial application and produces a narrower particle size distribution than the alternative wet granulation method of high shear mixing. However, like all granulation processes, it can be difficult to control and to scale up.

The particle size of granules in a tablet blend can influence flow behaviour in the tablet press, and also properties such as bulk density and compressibility

The particle size of granules in a tablet blend can influence, for example, flow behaviour in the tablet press, and also properties such as bulk density and compressibility. These properties go on to affect the CQAs of a tablet. As a result, particle size is a useful, easily-measured metric for understanding and scoping fluidised bed granulations, for controlling and tracking the trajectory of the process. However, the successful generation of particle size data in the lab, through scale-up and into production, relies on identifying the most relevant analytical technology for each step.

A modern size analysis strategy for granulation processes: The particle sizing techniques of laser diffraction and SFV both offer automated, rapid data acquisition and streamlined analysis with minimal manual intervention. These attributes make them attractive alternatives to sieve analysis in all instances, with the potential to reduce substantially the manual burden associated with gathering the information needed to implement QbD. The ease of use of offline laser diffraction systems and the resolution they provide make them especially helpful in the earlier stages of granulation development. It is at this point that laser diffraction can offer rapid, detailed scoping of the impact of different variables on the properties of the finished granules.

SFV, on the other hand, is a relatively inexpensive and effective choice for continuous inline monitoring during pilot scale studies and into commercial operation. At the pilot stage this enables efficient scoping of the design space, while continuous measurement supports effective process control within commercial production.

A productive, cost-efficient strategy for particle size analysis during fluidised bed granulation may therefore prove to be the application of laser diffraction analysis during early stage development, followed by a gradual switch to SFV during scale-up and into production. Following this strategy, especially from a starting point of using sieve analysis, relies on effective specification transfer. The following case study demonstrates the close agreement of results measured by all three sizing techniques and the consequent ease of specification transfer.

Techniques for granule sizing
Sieve analysis
Sieve analysis was one of the earliest established particle sizing techniques and remains in practice across industry because of its perceived low cost and simplicity. It is useful for the measurement of relatively coarse particles, up to several centimetres in size, but is more limited in the sub-100µm range where the strength of interparticle forces can result in agglomeration. This leads to poor reproducibility and/or sieve blocking. Using sieves to measure the feed to a granulation or assess its early progress may therefore be impractical. Wet sieving techniques can address this limitation, though at the cost of increasing measurement times.
From the perspective of day-to-day operation, sieve analysis is quite inefficient. Drawbacks include relatively long analysis times compared with other techniques, manually intensive measurement methods, and the need for rigorous maintenance to keep the sieves in good working order and sustain reproducibility. A further drawback is that sieve analysis is able to deliver only coarse data resolution, dividing a sample into only five to eight fractions. This means that subtle differences in granule size distribution, which could affect performance, may go undetected.
Laser diffraction
Laser diffraction is an ensemble particle sizing technique, which means it generates a result for the whole sample in one single measurement. This makes analysis times inherently short.
A sample, either wet or dry, passing through a collimated laser beam scatters light over a range of angles. Large particles scatter with high intensity at relatively narrow angles to the incident beam, while smaller particles produce a lower intensity signal but at wider angles. Laser diffraction analysers detect the angular dependence of scattered light intensity and calculate a particle size distribution for the sample from the resulting data, using the appropriate theory of light.
Laser diffraction offers precise resolution (up to 100 class sizes) across a very broad measurement range (0.01–3,500µm), making it equally suitable for the analysis of granulation feeds and many finished granules. High resolution capability provides precise detail on the particle size distribution, which gives valuable insight into exactly how the granulation is proceeding. From a practical perspective, laser diffraction systems are highly automated, so measurement is fast and requires minimal manual intervention.
Spatial Filter Velocimetry (SFV)
SFV is a number-based, chord length sizing method that generates a particle size distribution from measurements of individual particles. Particles falling through a laser beam interrupt the flow of light from the laser source to a linear detector array made up of optical fibres. Particle velocity is calculated from the sequential interruption of linearly neighbouring elements of the spatial filter detector triggering a burst signal, the frequency of which is directly proportional to particle velocity. Particle size is determined from secondary measurements of the length of time for which the particle blocks a single optical fibre. By measuring a large number of particles, SFV systems build up statistically valid data for the sample, from which various size parameters are calculated, depending on the needs of the user.
SFV measures across the range 50–6,000µm, making it especially suitable for tracking a granulation through to completion. Like laser diffraction, SFV can be applied to real-time measurement of a process in the form of an inline particle sizing probe, which is easily inserted into a pipe or vessel.
Relatively inexpensive and able to measure consistently and efficiently even in humid, sticky environments, SFV is a proven technology for the continuous monitoring of granulation processes through to the required end point.

Case study

Comparing the granule size data generated by sieve analysis, laser diffraction and SFV

A study was carried out to investigate the application of sieve analysis, laser diffraction and SFV to granulation process monitoring, to demonstrate whether the data produced using each technique were sufficiently similar to allow their interchangeable application.

A granulation was performed in a Glatt GPCG 1.1 fluid bed granulator using a model blend designed to reflect the properties of those used for tableting. The granules produced were then measured by sieve analysis (Electrolab sieve shaker), laser diffraction (Mastersizer 3000, Malvern Instruments) and SFV (Parsum IPP70, Malvern Instruments). The results for the sieve analysis are shown in Table 1 and indicate that the majority of the particles lie in the 1–1.18mm size range.

Table 1: Sieve data for a granule sample suggests that the bulk of the particle size distribution lies between 1.00–1.18mm
Sr. noDescription%w/w pellets
1Pellets retained on 14# mesh (1.4mm)0.00
2Pellets retained on 16# mesh (1.18mm)18.50
3Pellets retained on 18# mesh (1.00mm)68.40
4Pellets retained on 20# mesh (850ssm)12.80
5Pellets retained on 25# mesh (710ssm)0.00
6Pellets retained on 30# mesh (600ssm)0.00
7Pellets retained on 30# mesh (600ssm)0.00

For the laser diffraction measurements, granules were dispersed using Aero S, the integrated dry dispersion accessory of the Mastersizer 3000, which entrains particles in a stream of pressurised air. Dispersion ensures that particles are measured in an unagglomerated state, thereby improving the reproducibility and relevance of the laser diffraction data.

A major advantage of the Aero S is that it can handle the larger sample sizes required to reduce sampling errors, while maintaining rapid measurement times. This is particularly important for samples that are prone to segregation due to the presence of larger particles, such as may occur during granulation. The results of the laser diffraction analysis are shown in Figure 1. The results generated are highly repeatable with an RSD of 0.4% at the median (Dv50). The Dv50 is measured at 1,069µm, which is in line with the sieving data.

Figure 1: Particle size distribution data for the granules measured using laser diffraction is in line with the results from sieve analysis

Measurement numberDv10/µmDv50/µmDv90/µm
188510701253
291810731243
390910651236
Mean90410691244
Standard deviation1749
RSD1.9%0.4%0.7%

The SFV system employed also used an integrated dispersion cell to ensure the measurement of primary particle size. This technique produces number-based particle size distribution data which can be converted into volume-based distributions for direct comparison with laser diffraction data. The SFV results, both number- and volume-based, are shown in Table 2.

Table 2: Particle size distribution data for the granules measured using SFV is in line with both sieve analysis and laser diffraction
D10/µmD50/µmD90/µm
Number-based distribution2249851170
Volume based distribution88310951233

All three sets of data show close agreement for size measurement of the granules, with a difference of just 2% between the laser diffraction and SFV data. This suggests that size specifications can be relatively easily transferred from one technique to another. This in turn means that the sizing technique can be selected on the basis of the practicalities and requirements of measurement at different scales, and the resolution and quality of data required for effective product control.

In conclusion, laser diffraction and spatial filter velocimetry offer considerable advantages relative to sieve analysis for the measurement of granule size. Fast and highly automated, laser diffraction enables far higher resolution than sieving with considerably lower manual input, making it especially valuable in the earliest stages of granulation development. SFV systems on the other hand are a cost-efficient solution for inline sizing as a granulation proceeds, and provide timely measurement for process monitoring and control. Both techniques have the potential to save time and labour when gathering the data needed to implement QbD and develop a robust, well-controlled granulation process.

The data presented here provide reassurance that the specification transfer required to exploit the potential of both techniques can be easily implemented. Furthermore, it confirms that both technologies provide results that are closely comparable to those delivered by sieve analysis. It is clear that accessing the considerable benefits of switching from sieve analysis to laser diffraction and SFV can be both straightforward and rewarding.

Reference

1. N Moldavski et al. Exploring the benefits of real-time particle sizing – The optimization and scale-up of high shear granulation. Pharmaceutical Manufacturing, March 2015

Figure 1 Particle size distribution data for the granules measured using laser diffraction is in line with the results from sieve analysis

Companies