Developments in Powder Flow Testing

February 2, 2006
Pharmaceutical Technology, Pharmaceutical Technology-02-02-2006, Volume 30, Issue 2

The pharmaceutical industry's focus on process understanding, monitoring, and control is driving manufacturers to take greater steps toward identifying possible manufacturing bottlenecks earlier in the development process. For tablet, capsule, and excipient producers, such efforts include taking a closer look at the flow-ability of their powders.

The pharmaceutical industry's focus on process understanding, monitoring, and control is driving manufacturers to take greater steps toward identifying possible manufacturing bottlenecks earlier in the development process. For tablet, capsule, and excipient producers, such efforts include taking a closer look at the flow-ability of their powders.

There are only two truths when it comes to powder flow analysis: Powders are a complex material and no one single test method is the best. Everything else is debatable.

The US Pharmacopeia first started working toward standardizing the methods to measure powder flow with a survey and a "Stimuli to the Revision Process" article in Pharmacopeial Forum published in 1999 (1). Since then, a harmonized general information chapter on powder flow has been released for the more traditional methods (2) and sophisticated analyzers have been introduced and are slowly finding their way into pharmaceutical laboratories.

Quality and functionality

There is currently no real time, in situ method of measuring the flow of a powder during processing. "Although the community as a whole is making great strides in trying to understand more about what powder flow is, we're still a very long way off from being able to take a descriptor of the particles and predict what is going to happen without the intermediate, bench-scale tests that are out there right now," says James Prescott, senior consultant, Jenike & Johanson (Westford, MA, www.jenike.com).

Evaluating and predicting how a material will perform (i.e., its "functionality") is vastly different, and much more complex, than conducting typical characterization studies. "Functionality still gets people really worked up. I think physical test methods are a more acceptable concept that relates to functionality and performance. Powder flow is more of a functional or performance test than a physical test," says Gregory Amidon, research fellow at Pfizer (Kalamazoo, MI, www.pfizer.com) and chairman of USP's Excipients General Chapters Expert Working Group.

Part of the challenge is overcoming the industry's perception about what may constitute important processing parameters. As Prescott observes, "People think of the classic unit operation as deliberately doing something to the material, for example, getting rid of agglomerates or drying the material to a certain moisture. We think long and hard about the controls for that. But when you get into powder transfer, you don't want it to come out any different than how it went in because people think they are not doing anything to it, they don't think about all the parameters that are important to make sure that it works as intended. That is part of the reason why people don't think of functionality when it comes to powder flow. I think people want it to be simpler than it is, either out of being naïve or optimistic or hopeful or overwhelmed or any combination of these."

Variability challenges

Adding to the complexity are all of the sources of variability that can affect the accuracy of powder flow measurement. These can be categorized into three general areas: the physical properties of the particle (e.g., its shape, size, compressibility), the property of the bulk powder (e.g., size distribution, compaction), and the processing environment (e.g., storage, handling, humidity).

Just the physical nature of a powder makes it difficult to characterize. "Powder has the elastic properties of a solid, the compressibility properties of a gas, and can flow in all directions like a liquid," notes Reg Freeman (founder of Freeman Technologies, Worcestershire, UK, www.freemantech.co.uk). Consequently, powder has one set of properties when it's stagnant and another set of properties when it's moving. "There are quasi-static tests of powder properties and powder flow and there are dynamic tests—they are really quite different," says Amidon.

In addition, as Freeman observes, "No one quite knows what it is we are trying to measure. There are a lot of preconceptions. One is that a powder is 'good' or 'bad.' A powder, however, can completely transform. An aerated powder is completely different than one that is not, for example. Whether or not it is a good or a bad powder depends on the packing regime it's sitting in, for example."

The machinery that will be used will inevitably affect how the powder performs during processing. Changes in the speed of a tableting or encapsulation process or in the geometry and design of the hopper or bin, for example, can magnify discrepancies that were not detected during bench-top testing, even if there were no changes in the formulation (3).

In addition, personnel who run flowability tests may not have the breadth of manufacturing experience to understand how the result relates to what is happening on the manufacturing floor, and vice versa. The operators who deal with powder flow obstructions may not be the ones who are selecting the test methods to use to prevent those problems.

A new chapter

USP 29–NF 24, effective as of Jan. 1, is the first USP edition to include the recently harmonized General Information Chapter "‹1174› Powder Flow" (2).

The chapter provides general information only and gives the methods for characterizing powder flow that appeared most frequently in a review of more than 130 drug-manufacturing references. The chapter identifies experimental considerations for four methods: angle of repose, compressibility index or Hausner ratio, flow rate through an orifice, and shear-cell methods. And, it offers a "Scale of Flowability" table when applicable (see Table I). It provides recommendations regarding the standardization of the methods.

Table I: Scale of flowability*.

Testing methods: old and new

Some of the traditional flowability test methods described in the new USP chapter have been used for at least 40 years, and their advantages and limitations are well known (1). Each method reveals something different about the powder or its flowability, but more than one is needed. Says Amidon, "Although some would consider angle of repose, compressibility index, and flow through an orifice to be 'primitive,' there are enough data in the literature to indicate that they can be correlated with manufacturing experience and are therefore of some value."

Angle of repose. USP defines angle of repose as the "constant, three dimensional angle, relative to the horizontal base, assumed by a cone-like pile of material," which is formed when the powder is passed through a funnel-like container. The results rely heavily on the method used to form the cone. Two important variables are the height of the funnel and the diameter of the base (i.e., whether can be fixed or allowed to vary). The USP recommended procedure for conducting an angle of repose test includes having a fixed, vibration-free base with a retaining lip and maintaining a funnel height that is 2–4 cm from the top of the powder pile as it forms.

In addition to its simplicity, the angle of repose method has the advantage of having an associated general scale of flowability consistent with the classification reported by Carr (4) (see Table I). Disadvantages include that the material may undergo segregation, consolidation, or aeration as the cone forms. The method provides a general sense of powder flowability and thus is useful as an early indication of potential flowability problems. Says Amidon, "There are data showing a correlation between the angle of repose with weight variation off a tablet machine or a capsule machine."

Compressibility index. Compressibility index (and the closely related Hausner ratio), like angle of repose, is not an intrinsic property of the material. Unlike angle of repose, however, it is an indirect measure, relating a ratio (bulk–tapped volumes or bulk–tapped densities) with flowability. The chapter lists several experimental considerations and recommends a standard sample volume (250 mL), sample mass (100 g), and container shape (cylindrical).

Figure: Today´s sophisticated powder rheometers can measure how the energy needed to produce flow varies with air content and direct pressure consolidation levels.

Flow through an orifice. Empirically determining the flow rate through an orifice is useful only with free-flowing materials and not cohesive materials. (Prescott and Barnum define free flowing in (5)). The flow rate is generally measured as a mass rate or a volume rate of a powder flowing from a container and, as the chapter recognizes, can be in discrete increments or continuous. Like the previous two methods, flow rate through an orifice is not an intrinsic property of the powder. Unlike for the previous two methods, however, no general scale of flowability is provided because the rate is "critically dependent" on the method. The chapter lists experimental considerations and provides general recommendations for a procedure, including general guidelines for the dimensions of a cylindrical container.

Although flow tests provide information about what the flow is, it will not tell you why the flow is what it is. Says Prescott, "If it's more cohesive or more frictional, for example, the flow tests won't tell that it's because the material is more moist, or the particle size or shape changed, or something about its electrostatic charge changed. You can't determine what could have made the flow behavior change."

Shear-cell methods. To connect theoretical powder flow with real-world processing equipment such as bins and hoppers, manufacturers rely on shear-cell tests. There are many different methods, though they are variations on the same theme.

There are three basic designs: translational, also known as a Jenike shear cell, in which a cell is pushed in a linear direction to get the material to shear against itself in a linear manner; a rotational "plate" cell, also known the Peschel shear cell; and the rotational annular cell, also known as the Schulze shear cell. "These various shear-cell tester methods provide very similar information and very similar results. The real question is then, do you prefer a specific tester over another—something about the software or the convenience or the cell sizes or the support you might get from a company providing that tester. Those types of things become more important than the real differences in what the data are telling you," says Prescott.

According to the USP chapter, "A significant advantage of shear-cell methodology in general is a degree of experimental control. A general disadvantage, however, is that the methodology is rather time-consuming and requires significant amounts of material and a well-trained operator." Because of the diversity of the methodologies, the chapter does not provide specific recommendations, saying only that "It is recommended that the results of powder flow characterization using shear-cell methodology include complete descriptions of equipment and methodology used."

One of the strengths of shear-cell tests is that once a test is conducted, it can help predict the bin geometry, shape, angles, surface finish, and opening sizes that one would need to prevent flow problems. "Many other test methods simply rank materials and don't provide specific engineering guidance on what to do to prevent problems. If you get a certain angle of repose but you can't change the formulation, what is the next step? I think that is one of the weaknesses with qualitative methods," says Prescott.

Shear-cell methods are especially useful when the formulation cannot be changed. Says Prescott, "That's the strength of shear testing and looking at the data in that light. Flow is tied inherently to the nature of the formulation or the blend as well as the equipment that is handling that blend. If you can't change something about one, then you need to start thinking about changing something about the other. You always have to mate the two together."

Shear-cell methods also have some disadvantages. "The problem with the shear-cell method is that you can generate a lot of information but it is difficult to correlate that information with the actual flowability of a powder," says Jian-Xin Li, Pharmaceutical Projects Manager, at FMC Biopolymer (Princeton, NJ, www.fmcbiopolymer.com).

Of the four methods discussed in the USP chapter, shear-cell methodology is most complex and will need expanding. Says Amidon, "This chapter predates the developments for automated shear-cell methodologies. It's missing an assessment and a discussion of the improvements in the technology over the past few years to have commercially available automated or semiautomated shear-cell methods that are reproducible and useful. That's where I see this chapter being expanded."

Vibration. Methods that include vibrating mechanisms—such as vibrating spatula or hopper—are used in the industry and are commercially available. USP does not discuss these methods, but it does mention that the use of a vibrator to facilitate flow from a container "appears to complicate the interpretation of the results."

Avalanche test. The avalanche test is a dynamic method. Unlike other tests, this method does not put the powder under stress, so it is best used for powders that are free flowing. During manufacturing, this free-flowing condition occurs during blending, conveying, tableting, capsulating, and packaging.

Although the avalanche technology was established about 15 years ago, researchers are finding new approaches for conducting avalanche tests (6). The avalanche time is captured photoelectrically and a phase-space attractor plot is generated. The mean time to avalanche (the attractor or centroid) of the pattern reveals a flowability "index," and the scatter is an indication of the powder's cohesivity.

"The biggest advantage to the avalanche test versus any other method is that within 20 minutes you can get several hundred tests completed, which increases the statistical significance and lowers the deviation," says Brent Kiser, senior applications engineer at TSI Inc. (Shoreview, MN, www.tsi.com).

Avalanche testing also is not discussed in the USP chapter, though studies have been conducted to help bring some standardization in testing (7) and the method has been useful in other industries (8). Studies also have shown that avalanche data correlates with tablet variability for direct-compressed tablets (9) and have demonstrated the value of operator observations during testing (10).

Powder rheometers. Powder rheometers are relatively new testing instruments that can sensitively measure how flowability changes under a wide range of processing conditions, including various speeds, levels of entrapped air, and degrees of attrition. Powder rheometers also can condition the sample before testing, reducing the variability produced by differences in storage and handling.

These instruments automatically rotate a blade through a cylindrical column of powder, measuring the energy or force it needs to move through the sample and relating these measurements to various characteristics of the powder.

Says Freeman, "With our flowability measurements, we want to determine the factors that might relate to these processes. How easily does the powder flow? Does the powder move when it is not consolidated, and how does consolidation affect it? Will the presence of air affect it?" These instruments are versatile because flowability is only one of several applications and manufacturers continue to offer new options in line with customer needs. Studies have shown the usefulness of a powder rheometer to evaluate the flow properties of powder granules before and after lubrication (11) and how the presence of air in a powder affects the rheology of a powders (12).

Pharmaceutical users most often seek tight blade controls and tight manufacturing controls, so that the answer they get in one facility is the same as the answer they get in another. In most cases, they also want 21 CFR Part 11-compliant software. "They may not want to use it, but they want to know they have the ability to use it," says Marc Johnson, president of Texture Technologies Corp. (Hamilton, MA, www.texturetechnologies.com).

Users also want to be able to customize the instrument to highlight the factors most important to them, so the software must be flexible. Instruments differ mainly in their software, blade design, and how the test is evaluated. For example, one analyzer may measure both the force and the torque (model FT4, Freeman Technologies) while another (TA.XT2i, Texture Technologies Corp. and Stable Micro Systems) measures the force and not the torque.

These instruments don't necessarily replace other methods. Says Johnson,"Of all the powder analyzers that we've sold, I don't think anybody doesn't also do a Jenike shear cell if they feel they need to."

Method selection

No single and simple test method can adequately characterize the flow properties of pharmaceutical powders. Instead, most scientists advocate using multiple standardized test methods to characterize various aspects of powder flow. Some tests yield information about the physical properties to predict what may happen during manufacture, and others serve as a ranking or correlation tool.

The challenge then becomes ensuring each test is conducted correctly to obtain meaningful results. Says Johnson, "There are various techniques out there that are wonderfully precise, but the problems tend to be that people don't know yet how to use them. They don't know under what circumstances they need to conduct avalanche testing or a Jenike shear test. So there are a lot of training issues that need to be resolved."

Numerous studies have shown—or have attempted to show—a correlation among the various methods (13, 14). Says Prescott, "I think people hope there would be a single test that would just say that everything is going to be okay or not—an angle of repose that is better than x or a density or Hausner ratio that is better than y—as a cut-off point. Our experience is that there are many properties that come into play and, more important, those properties need to be looked at with respect to the very specific application at hand."

New technology for measuring flow may not necessarily replace the more conventional methods. For example, excipient makers are more likely to provide particle size data and other physical properties about their products than they are to report on a powder flowability test. "We see particle size, shape, and density, and particle size distribution as part of the certificate of analysis because those are critical parameters that will affect the flow of powders," says Li.

Amidon agrees, "The standard rationale for public standards and the existence of USP is to publish meaningful tests for strength, identity, purity, and quality (15). Strength, identity, and purity are chemical properties. Quality really gets at the heart of physical properties that relate to excipient performance. More often than not, what determines excipient quality is its physical properties. While in the end what we're interested in is powder flow, at least when it comes to excipients, we tend to worry more about the physical properties and assume the flow properties will follow."

Modeling flow

Manufacturers are seeking methods that would provide some early indication of manufacturing problems, even at the research-and-development and pre-formulation stages. As Johnson observes, "One of the trends we believe will continue is the desire for people to model flowability from very small powder samples," because supplies of API or drug product are scarce. "Because the flowability of very small amounts of powder gets distorted by the sidewalls and by the number of surfaces the powder is in contact with, methods such as angle of repose, shear cells, avalanche methods, and funnels are not suitable or effective. Many remain highly variable and do not generate operator-independent results."

Many people have found that the flowability of their powders can be modeled from stress relaxation tests. Models that connect interparticle bonding to powder cohesion and flow may someday produce tests that predict flowability (16).

Other studies have brought artificial neural network models (17) to bear on the effects of bulk, tapped, and particle densities, particle size, and shape on the flow rate.

Process modelers also are applying statistical tools. Says Lynn Torbeck, statistician and member of USP's Statistics Committee, "One of the things the USP Statistics Committee currently plans to work on this cycle is providing statistical direction in the area of bulk sampling of powders using references by Pierre Gy and others. We're talking about applying current theories in bulk sampling to pharmaceuticals, recognizing that there will have to be assumptions or approximations."

Future goals include being able to evaluate the flow of powders in-line during manufacture, thereby monitoring the process and controlling variability. Says Prescott, "We always had the mantra of: Know your process, know your material. Know what equipment is being used to handle your material, and know what you expect for properties of that and for flow behavior. That makes all the difference in the world. It's really getting a full understanding of what's there, which gets back to functionality and gets back to the directives from FDA in terms of PAT and design space goals."

Inching toward progress

The complexity and challenges of measuring such an abstract characteristic as powder flowability also leaves plenty of room for projections about where the science and industry is heading. "I see powder flow becoming much more practical. What are the problems and how do you solve them? You can have 10 different companies making aspirin, but their concerns will be different, their equipment will be different, or they may care about very different aspects of powder flow. I see the solutions following the problems and being much more empirical," says Johnson.

Others envision using these empirical results to create databases that would help to predict flow behavior, especially as the data from these tests improve and process understanding increases. Says Freeman, "We want to quantify what is going on. That is now very possible, because now we can create databases of powder flow properties. We never had that in the past. These databases can help to determine which parameters would be most important in processability." Says Li, "If a company has PAT, monitoring the flow is certainly part of the picture. Adjusting the process based on the flowability result is part of PAT."

As the objectives of PAT take their course with respect to process understanding, as the industry will move forward in understanding its processes, so too will the understanding of how their materials behave under these processing conditions. Says Freeman, "Inevitably if there are no ways of determining in a precise way what is causing the weight variability of the active in these tablets, then there is no great pressure to provide better data. It's changing slowly. Questions are starting to be asked, and especially with PAT, people now want to understand the process. There is no simple answer and nobody fully understands what is going on yet, but we're all working towards achieving that enlightenment one day. Hopefully, eventually, we'll all join up."

References

1. G.E. Amidon et al., "Physical Test Methods for Powder Flow Characterization of Pharmaceutical Materials: A Review of Methods," Pharm. Forum, 25 (3), 8298–8308.

2. "‹1174› Powder Flow" in USP 29–NF 24 (US Pharmacopeial Convention, Rockville, MD), p. 3017.

3. J. Prescott et al. "Comparison of Flow Properties of Granules Prepared by Top-Drive and Bottom-Drive High-Shear Granulators" poster presented at the AAPS annual meeting 2004.

4. R.L. Carr, "Evaluating Flow Properties of Solids," Chem Eng. 72, 163–168 (1965).

5. J.K. Prescott and R.A. Barnum, "Powder Flow," Pharm Technol. 24 (10), 60–84 (2000)

6. F. Lavoie, L. Cartilier, and R. Thibert, "New Methods Characterizing Avalanche Behavior to Determine Powder Flow," Pharm. Res. 19 (6), 887–893 (2002).

7. B.C. Hancock et al., "Development of a Robust Procedure for Assessing Powder Flow Using a Commercial Avalanche Testing Instrument," J. Pharma. Biomed. Analysis 335 (5), 979–990 (2004).

8. S. Rastogi and G.E. Klingzing, "Characterizing the Rheology of Powders by Studying Dynamic Avalanching of the Powder," Part. Sys. Character. 11 (6), 453–456 (1996).

9. K.M. Lusvardi et al., "Powder Flow Characteristics of Directly Compressible Hydroxypropylcellulose Modified Release Matrix Systems," PTR-020, retrieved Nov. 18, 2005, www.aqualon.com.

10. Y.S.L. Lee et al., "Development of a Dual Approach to Assess Powder Flow from Avalanching Behavior," AAPS PharmSciTech 1 (3), article 21 (2000).

11 C.V. Navaneethan, S. Missaghi, and R. Fassihi, "Application of Powder Rheometer to Determine Powder Flow Properties and Lubrication Efficiency of Pharmaceutical Particulate Systems," AAPS PharmSciTech 6 (3), E398–E404 (2005).

12. R. Freeman, "The Importance of Air Content in the Rheology of Powders: An Empirical Study," Am. Lab News, Nov. 2004, accessed Jan. 4, 2005, www.freemantech.co.uk.

13. R. Freeman, "Measuring the Flow Properties of Consolidated, Conditioned, and Aerated Powders—A Comparative Study Using a Powder Rheometer and a Rotational Shear Cell," Particulate Systems Analysis 2005, obtained from author.

14. M.K. Taylor et al., "Composite Method to Quantify Powder Flow as a Screening Method in Early Tablet or Capsule Formulation Development," AAPS PharmSciTech 1 (3), article 18 (2000), retrieved from www.pharmscitech.com.

15. USP Constitution and Bylaws, Section 3.

16 D. Bika et al., "Strength and Morphology of Solid Bridges in Dry Granules of Pharmaceutical Powders," Powder Technol. 150 (2), 104–116 (2004).

17. K. Kachrimanis, V. Karamyan, and S. Malamataris, "Artificial Neural Networks and Modeling of Powder Flow," Int. J. Pharma. 250 (1), 13–23 (2003).