Physical Quality Specifications Lead to Success

Physical Quality Specifications Lead to Success

By Guy Samburski, Senior Director of Technology R&D

Whether a finished drug product is in the form of a tablet, capsule or sophisticated delivery system, the API’s physical properties play a key role in determining whether that particular dosage form will perform effectively.

Particle size distribution, or PSD, is probably the most important physical property of an API. It is a critical parameter in the manufacturing process and determines how effective a dosage form’s performance will be. In order to meet the specifications, the API provider must possess expertise in crystallization and size reduction processes. At TAPI, our experience has taught us that it is essential to collaborate closely with pharmaceutical manufacturers and to set the right specifications to ensure the API is manufactured flawlessly and that there is a consistent supply.

physical specs by TAPI


The approach to setting PSD specifications is different than setting chemical specifications in few ways:

  • Chemical quality primarily refers to the purity of the API. Specifications usually define an acceptable upper limit of a certain impurity (i.e., percentage of impurity, residual solvent, etc.). In PSD, a range is defined and particles that are too large or too small are not acceptable. Therefore, the specification will be bi-sided.
  • Chemical quality is a relatively universal concept. The limits will often be well defined in pharmacopoeias, official quality guidelines and more, making it fairly easy to obtain a specification limit that is accepted by many authorities and manufacturers. However, with Physical quality is client-specific. Every finished drug product manufacturer may have its own preferred specification based on its formulation and manufacturing process.
  • There are no universal methods for PSD measurement methods. Pharmacopoeias do not define such methods, and different methods such as sieving or laser diffraction microscopy will deliver different results. Even different instruments working on the same principle, such as different models of laser diffraction instruments, will deliver different results. In addition, each method is product-specific.


As a result, the API provider should adopt a case-by-case approach and work closely with the pharmaceutical manufacturer to ensure specific needs are met.


The Details are Essential

During dissolution, particles of median size will dominate the rate of dissolution. The dissolution curve may have a “tail” of longer dissolution times that is dominated by the last few coarse particles to dissolve. For that reason, it is necessary to control the content of both average and coarse particles.

When it comes to the dosage form’s manufacturability, other properties are dominant. Material needs to be well-flowing, non-sticky, and must easily blend with excipients. These factors are primarily dictated by the fine particles in the powder. However to manufacture a viable product, fine particles must be limited. Dissolution works better with fine particles while manufacturability favors coarser particles. A well-defined specification that balances these requirements is the key to production robustness.

It is important to remember that the different estimators for PSD are not independent. If the API is milled in order to reduce d(0.9), the d(0.5) will decrease as well.  

In addition to PSD, other physical properties play an important role in pharmaceutical manufacturing, notably Bulk Density and Tapped Density. The difference between the two values is an indication for the flow and compressibility properties of the material. In almost all cases, these values will need to be as high as possible and close to each other to easily process well-flowing material.


Statistics Rule!

APIs are powders that contain a multitude of particles, each with a different size. The population of the particles is commonly called a statistical point estimator. For example, the term d(0.5) is the population’s median particle size and is a statistical point indicator. One of the most important concepts that applies to these terms is the central limit theorem. This fundamental statistical rule implies that statistical point estimators will always be distributed normally, regardless of the distribution of the population they describe. This concept can be easily illustrated by the following simple dice-throwing game:

Physical specs example


When a single die is thrown, the distribution is uniform. The chances of obtaining any value from one to six are equal. However, the arithmetic mean of many rounds with several dice will converge in a normal (Gaussian) distribution with a given average and standard deviation. So although API manufacturing is a well-controlled process, the median PSD [d(0.5)] of many batches will always have a certain normal distribution with a given average and standard deviation. Tight control over the crystallization parameter and mill settings may reduce the standard deviation, but some variance will always be there. The same applies to d(0.9), d(0.1), etc. Therefore, a product specification is best defined when it is based on statistical data from many batches. TAPI scientists have developed a methodology to generate specifications with statistical significance based on a wide database of many commercial batches, as well as on a limited number of lab and pilot batches for new products under consideration.

Physical target API statistics


Trends in Regulatory Requirements

Today, most health authorities recognize that PSD and other bulk physical properties should be defined in specifications. The current trends include:

  • Requiring three-tiered specifications. For example, requiring limits on d(0.1), d(0.5) and d(0.9).
  • The limits may be one-sided. In a few cases, as with the Canadian authorities, a bi-sided limit is required for the median PSD [d(0.5)].
  • Other physical properties should be defined if they are found to be a Critical Quality Attribute (CQA). For example, Bulk Density and Tapped Density will often be considered as CQAs.


Wide or Narrow?

The range of bi-sided specifications should be carefully selected. Ranges that are too wide may cause the final dosage form to fail dissolution. Therefore, the limits should be selected with some safety margins that are evaluated during the development of the formulation and the biostudy. On the other hand, ranges that are too narrow compared to the normal distribution of the provider’s API batches may make it difficult to consistently manufacture material that complies with the specifications. To better control the PSD, the vendor may need to upgrade technology which will impact supply times and costs. While the specifications should definitely ensure the dosage form consistently meets the standards, they should take into account the manufacturer’s ability to conform in a cost-effective manner.

recommended formats for PSD specification - TAPI


Recommended Format for PSD Specifications

  • Three tiers: d(0.1), d(0.5), d(0.9)
  • Control of coarse particles: Upper limit for d(0.9): d(0.9) NMT XX μm. This limit will ensure dissolution.
  • Control of average particles: Bi-sided range for d(0.5): YY μm < d(0.5) < ZZ μm. This limit will ensure consistency.
  • Control of fine particles: Limit the bottom range for d(0.1): d(0.1) NLT μm. This will ensure manufacturability.
  • Bulk and Tapped Density: Set a bottom limit for these values.


Accurate PSD Specifications Begin with Reliable Data

For commercial products, TAPI maintains a statistical database of physical properties from at least 10 batches on a commercial scale. In many cases additional data is available from multiple sites, historical trends, etc. This data is used to propose specifications and recommend analytical methods to customers. If the customer uses a different PSD method, both should work together to agree on the specification limits.


Creating specifications for a new product is more challenging because of the absence of a statistically significant database. TAPI uses statistical tools to forecast the future variance of production batches based on limited commercial data, with additional data from smaller-scale manufacturing. Over time, this process has been proven to provide accurate limits for specifications. As with commercial products, it is essential that the API provider and customer align their PSD analytical methods.


It takes a strong working relationship between the API provider and finished drug product manufacturer to produce and deliver high quality APIs. Careful planning, open communication and close collaboration with a trusted partner will ensure that you receive a consistent supply of superior products – the key to your success in today’s marketplace.


Guy Samburski is Senior Director of Technology R&D for Teva Pharmaceutical Industries, Ltd. with over 20 years of experience supporting pharmaceutical manufacturers with formulation, physical properties and other technological issues.


>>>Contact us to learn how TAPI’s team of experts can meet your specialized needs