Optimizing Tableting Processes with Quality by Design

PharmTech: In your experience, What attributes besides particle-size distribution must be considered for tableting processes?

Freeman (Freeman Tecnology): As the industry focuses on better manufacturing efficiency, there is greater interest in identifying powder properties that directly influence tableting in-process performance and final product quality. Particle-size distribution is a critical primary particle characteristic of powders, but it is only one of many variables that impact bulk powder properties, which in turn dictate in-process behavior and product quality. Bulk property measurements can be an efficient way of accelerating and supporting process optimization studies because they quantify the net effect of all primary particle properties (e.g., size, shape, texture, surface energy and porosity), whether these can be measured directly or not. Furthermore, even if all primary particle properties that influence in-process behavior could be measured, the mathematical relationship between bulk powder behavior and particle characteristics remains elusive and highly complex. Hence, the most effective way forward is to measure process relevant characteristics of the bulk powder.

Tim Freeman

Tablet production can be divided into at least four discrete processes: discharge from the hopper; flow into and through the feedframe; die filling; and compression. Each of these processes subjects the powder to a specific set of environmental conditions (e.g., flow rates, stresses, and equipment surface properties), making different bulk properties more relevant at different stages. I would highlight the following as especially valuable:

• Dynamic flow properties (including Basic Flowability Energy, Specific Energy, Aerated Energy, and Flow Rate Index): to optimise the flow regime in the feedframe and the efficiency of die filling, to investigate the effect of paddle geometry, to assess the likelihood of attrition, segregation and agglomeration.

• Shear properties: for optimising flow from the feed hopper, where shear properties of powder–powder and powder–wall are important.

• Permeability and compressibility: for assessing how easily the powder can transmit air and the impact of compression on the powder. Both characteristics are important during the filling and compression steps.

WALTER B. MCKENZIE/PHOTODISC/GETTY IMAGES; COMPOSITING BY DAN WARD

Levoguer (Malvern Instruments): Success in tableting does indeed depend on many factors. It is important, for example, to control the flowability and compressibility of the tableting blend, as well as any tendency towards segregation, to ensure the production of uniform tablets at the required rate. Particle size and particlesize distribution are recognised as critical material attributes because they are known to directly impact these properties, as well as others such as solubility and bioavailability, which may define clinical efficacy as highlighted in ICH Q6A.

Carl Levoguer

As analytical techniques evolve, however, it is becoming easier to identify other parameters that also impact behavior in the tablet press. Here I would highlight particle shape, a parameter that, like particle size, is known to affect powder flowability and segregation. In the past, shape information was gathered by microscopy, but the advent of automated imaging has made it much faster and easier to access statistically relevant data. Such information forms a foundation for scientific investigation of the impact of shape and supports the development of more successful tableting blends.

PharmTech: How are quality-by-design (QbD) approaches changing the way that tableting and granulation processes are viewed?

Freeman (Freeman Technology): QbD calls for product quality to be 'designed in' rather than tested for in postproduction. It requires a detailed understanding of all the factors that can impact product quality and clinical efficacy, including those related to the materials employed and the process itself. Traditionally, it has been assumed that raw materials and intermediates can be suitably qualified and the process can be fixed, resulting in a consistent high-quality product. However, this is only achieved by knowing what material properties need to be qualified. While particlesize distribution is important, there are many other particle properties that rarely feature in the specification, but that can be as influential as particle size, such as particle shape and particlesurface roughness. Excluding these properties from the quality specification allows variation in raw materials to go undetected, resulting in variable in-process performance and product quality. Adopting a QbD approach requires an acceptance that raw materials are likely to vary batch to batch, while simultaneously demonstrating a good grasp of how to configure the process settings within the 'control space' to accommodate the unavoidable variation in material properties, and ultimately achieve consistent product with the desired attributes.

Considering a granulation process as an example, this might conventionally be defined in the following terms: process for X minutes at an impeller speed of Y rpm, whilst adding Z% of water at a consistent addition rate. Processing conditions are essentially fixed and applied to each new batch of feed. This means that there is little flexibility to respond to variability arising from any source, such as a new batch of excipient or inadequate control of an upstream operation, for example. Furthermore, problems are usually detected only when granulation is complete.

QbD places emphasis on controlling process output, rather than the fixed definition of operating conditions. For granulation, the process definition might change to: manipulate impeller speed, amount of water, and/or processing time, to produce granules with these specific properties. Adopting this approach, however, relies on being able to identify those specific properties—the criteria for success—and also learning how to control them.

In the same way, in tableting, a QbD approach would focus on the defining characteristics of the finished product, such as content uniformity and dissolution or disintegration properties. Process development then works back from that point, identifying all the factors that influence these properties.

Levoguer (Malvern Instruments): Successful implementation of QbD relies on understanding both the process and product in detail. The focus is on fully evaluating the impact of all variables that influence product quality, and learning how to control them effectively, rather than just identifying a manufacturing route that works. QbD extends through to control of the commercial process so it serves to highlight areas where real-time monitoring can be beneficially applied to meet processing targets.

One important feature of particlesize analysis is that, unlike many analytical techniques, it is already a proven technology for real-time plant monitoring. In granulation processes, for example, both in-line probes based on spatial particle velocimetry and on-line laser diffraction particle size analysers are regularly used for real-time measurement. Both enable the continuous tracking of particle size growth during the granulation process towards an established endpoint.

Endpoint detection is a notoriously difficult aspect of granulation so this ability to continuously monitor particle size is extremely useful when manufacturing to meet a defined output, as advocated by QbD. In addition, however, real-time measurement is extremely valuable during design space scoping studies because it enables rapid and reliable assessment of the impact of a change in operating conditions. Continuous particle-size measurement can therefore accelerate and improve the process development studies associated with QbD.

PharmTech: What key challenges continue to exist with regards to understanding particle attributes in a tableting and granulation process?

Freeman (Freeman Technology): The bulk properties that define processability depend on a wide array of particle attributes, such as particle size and shape, roughness, surface charge, density and porosity. Learning how to control tableting and granulation processes relies, in part, on understanding the relationships between particle attributes and bulk powder properties.

This is an area of specific interest to Freeman Technology and we have been involved in a number of experimental studies, with industrial partners, to investigate, for example, the influence of particle size and shape, and of surface charge, on powder flowability, shear properties and bulk parameters, such as compressibility and permeability (1, 2).

Levoguer (Malvern Instruments): Because QbD places emphasis on thoroughly understanding the impact of all processing variables, it may call for information that is not easily accessed using conventional testing methods. As a result, the implementation of QbD is encouraging the pharmaceutical industry to adopt new analytical technologies as they become available. One such technology is morphologically directed imaging, which can combine imaging technology with spectroscopy, such as Raman, to provide chemical identification alongside size and shape measurement. It allows different particles in a dispersed sample, often initially screened on the basis of size or shape, to be reliably identified as specific chemical entities.

A conventional way to assay a tablet is to dissolve it and carry out high-performance liquid chromatography analysis. This gives an averaged measure of the concentration of the active that can be used to assess dose consistency, but it provides no information about the size of discrete active particles that are delivered to the body as the tablet disintegrates. In contrast, applying morphologically directed imaging to a disintegrated tablet sample allows differently sized elements of the resulting powder to be precisely identified as active or excipient. This not only generates useful information for engineering sophisticated drug delivery profiles, but also provides evidence to support claims of bioequivalence for a generic product.

PharmTech: What challenges exist with regard to understanding bulk-powder attributes?

Freeman (Freeman Technology): I think it is reasonable to say that the pharmaceutical industry's ability to understand how bulk powder properties impact process behavior has been constrained by a lack of reliable bulk powder property data. The reproducible measurement of defining powder characteristics, such as flowability, has long been a goal, but the results have been mixed. Traditional techniques, such as flow through an orifice and tapped density methods, are not ideal for the extended, detailed experimental work required to support QbD. Shear-cell measurements are ideal for understanding flow in hoppers, but are less useful for understanding lower stress processes, such as mixing, filling and aerosolization. Here, different measurement techniques are required.

Powder testing has developed considerably in the past decade, including the introduction of dynamic testing. Dynamic characterization reproducibly and directly measures powder flowability, for conditioned powders and for those that are consolidated or aerated, thereby generating reliable and valuable information for process development. Used in combination with bulk and shear property measurement, dynamic testing enables the kind of multifaceted powder characterization required to fully rationalise in-process behavior.

With these techniques in place, it is now possible to develop a detailed understanding of the way bulk properties influence tableting, granulation and many other frequently employed unit operations. This type of knowledge development remains a work in progress, but the goalposts shift too. Faster tableting speeds are one example, but the long-term objectives of continuous production in integrated manufacturing suites adds another layer of complexity, requiring testing strategies that provide the deepest and most comprehensive information.

Levoguer (Malvern Instruments): Generally speaking, my area of expertise is the measurement of particle properties rather than bulk powder attributes of the material, such as flowability and permeability. However, I've recently observed that some of our more experienced laser diffraction customers within the pharmaceutical sector are now using particle size analyzers to directly access information about the cohesivity of the sample because this is a useful property for tablet blend optimization.

Sample dispersion is an essential element of laser diffraction particle-size analysis. In dry measurement, dispersion conditions are established by conducting a pressure titration, which involves gradually increasing the pressure of the dispersing gas until a steady particle size is measured and/or the results obtained are identical to those generated with wet dispersion. Agitation and sonication are applied to achieve wet dispersion. In either case, the development process used to develop a robust method yields parameters that effectively quantify the strength of particle-particle interactions within the powder sample.

In state-of-the-art laser diffraction systems, dispersion conditions can be very precisely controlled and this measure of cohesive strength can therefore be sufficiently sensitive to give useful information for tablet blend optimization.

PharmTech: What technological gaps exist among the industry for fully understanding the material properties of powders used in a tableting process?

Freeman (Freeman Technology): Particles exhibit a wide range of mechanical and chemical properties. Several of these are commonly recognized as being highly influential with respect to in-process performance and product quality, and are often well defined as part of the specification. However, other properties, which may be even more important in some instances, such as particle shape, roughness, porosity and stiffness, to name just a few, are rarely considered in a specification because they are either too difficult to measure directly or can't be measured at all. Nevertheless, that doesn't exclude them from influencing the overall bulk material properties and, consequently, the in-process performance of the powder and the final product quality.

Direct measurement of many of these properties will only be achieved with further technological advances, but they will be required if the goal of "mathematical QbD" (accurately predicting tablet or granule properties from a knowledge of a range of particle properties) is to be achieved. In the interim, recent advances in bulk powder measurement techniques, simulating the conditions powders observe in the process environment and quantifying their response to those conditions, are likely to provide the best opportunity for understanding the relationship between bulk material properties, process parameters and final product attributes.

Levoguer (Malvernns Itruments): As I've previously suggested, one of the reasons why our customers measure parameters, such as particle size and shape, is because they are known to correlate with processability issues such as flowability, compressibility and the likelihood of blend segregation. These relationships are widely recognized, but my understanding is that they are not yet quantified in ways that help to establish optimal bulk powder properties. I would, therefore, point to lack of knowledge in the area of correlating particle and bulk powder properties as a gap that needs to be filled in the future.

As I noted previously, the advancement of automated imaging has made it much easier to access the shape data needed to extend our understanding in this important area and we have done some work in this area ourselves with Freeman Technology looking at the effect of particle size and shape on the bulk properties of lactose (1). This is an interesting area of research and one that I'm sure will receive further attention from the pharmaceutical industry as it works to extend understanding of how to make powders behave as required.

PharmTech: Can you offer any best practices for those beginning to apply a QbD approach to tableting and granulation?

Freeman (Freeman Technology): QbD relies heavily on engineering an optimized, well-understood process. It is therefore important, from the outset, to work out how to gather analytical data that will accurately reflect process performance. Effective powder handling is central to the success of tableting, granulation and a wide range of other pharmaceutical unit operations. Appropriate powder characterization techniques are, therefore, an essential prerequisite.

The number of powder testing techniques available reflects both the importance of such testing and its difficulties. When choosing which techniques to apply for QbD studies, I would suggest assessing against a number of criteria including:

• reproducibility and sensitivity

• process relevance

• ease of use.

Levoguer (Malvern Instruments): One of the biggest challenges for those applying QbD is how to access and gather the necessary information. The full implementation of QbD demands a comprehensive understanding of process and product, and the identification of an effective control strategy for the manufacturing process. Choice of analytical instrumentation is therefore crucial.

With well-established techniques such as laser diffraction particle-size measurement, customers can rightly expect the highest levels of automation and analytical productivity. Some systems can extend the efficiencies of dry measurement to more samples and combine rapid measurement times with assured data quality, to push analytical productivity to high levels for all users.

Of equal importance, however, are continuous laser diffraction particle size analyzers that offer real-time measurement for pilot-scale studies and commercial plant monitoring and control. These systems can significantly accelerate QbD studies. Running a pilot plant with real-time monitoring in place enables consistent control at the experimental conditions of interest and makes the impact of changes in operating variables instantly obvious.

For some types of analysis, the technology is newer, but it is vital to recognize what can now be achieved. Returning to the example of morphologically directed imaging, these systems involve considerable investment but can deliver significant value over the long term. Being able to measure not just size and shape but also the distribution of different chemical species within a dispersed sample, such as a disintegrated tablet, can be invaluable when trying to really understand how the process works and how to optimize it.

References

1. X. Fu et al., Particuology 10 (2), 203–208 (2012).

2. J. Khoo et al., "Use of Surface Energy Heterogeneity to Relate the Effect of Surface Modification to Powder Properties" (Freeman Technology website, 2012), www.freemantech.co.uk/goto.php?link=ART_POSTER_02, accessed Apr. 16, 2012.