ESEM provides information critical to understanding dissolution behavior of nanoscale particles.
Distinguishing between size and shape requires image-based analysis, but the trend toward smaller particles has pushed the requirements beyond the capabilities of light microscopy. SEMs have the needed resolution but their use has been limited by the requirements they place on the sample and its environment. ESEM relaxes these requirements and provides an attractive alternative for particle size and shape analysis when developing new methods or interpreting light scattering data to control existing processes.
The molecular size and complexity of drugs show a tendency to grow. Larger molecules are often less soluble. In order to maintain bioavailabilty, drug manufacturers have to balance increased size and decreased solubility at the molecular level with decreased size and increased solubility at the particulate level. It is not unusual for modern compounds to exhibit particle sizes measured in nanometers.
Light scattering techniques for particle size measurements fa]l into two main categories, often referred to as static and dynamic. Static measurements rely upon the inverse relationship between particle size and the angle at which the particle scatters incident light. Detectors measure intensity as a function of scattering angle and derive particle size based on Mie scattering theory. Static techniques can measure particles down to about a hundred nanometers in size.
Smaller particles require dynamic techniques (also known as photon correlation) that examine the effect of Brownian motion on light scattering. As might be expected, static scattering intensity varies about some mean value, however, the variations are not discontinuous but temporally correlated and reflect the continuous Brownian motion of the particles. The speed of motion is a function of particle size, specifically, the hydrodynamic radius of the particle. Theoretically, dynamic scattering can measure particle sizes down to the single nanometer range.
Both static and dynamic scattering assume particles have a spherical shape. The values they report are nominally particle size however they are in fact some collapsed function of size and shape. This makes them ideal for process control applications where fast measurement is necessary for effective control in a production environment. Since the measurement contains shape information, it is sensitive to shape changes. Typically these appear as changes in the shape of the particle size distribution. This is sufficient to alert the process engineer about changes but not to understand the nature of the change.
A detailed understanding of shape requires image-based analysis. For larger particle sizes this usually involves the use of optical microscopy. Automated image analysis software performs most of the lower level work and can report a large number of statistically valid shape characteristics. However, much of the power of image-based analysis derives from its qualitative capabilities. For instance, many drug compounds are crystalline. An experienced microscopist may quickly diagnose process problems from changes in crystalline structure that cannot be seen in scattering data.
The resolution of a light microscope is limited by the wavelength of visible light to about 0.5 micrometers. Practically, this limits the use of light microscopy for shape analysis to particles several micrometers or more in size. Smaller particles require the use of SEM. An SEM forms an image of the sample by moving a finely focused beam of electrons over the surface and mapping, point by point, the signals generated by interactions between electrons in the beam and atoms in the sample. SEM resolution is typically a few nanometers. The "depth of focus" is another significant advantage of SEM and allows imaging of particles over a large size range.
In addition to high resolution, SEMs also exhibit deep focal fields. In particle analysis applications, their ability to simultaneously image small particles near the substrate surface and large particles that may extend well above that surface a]lows them to accurately evaluate size and shape for samples with broad size distributions.
Generally SEMs impose two kinds of constraints. First, to prevent scattering of the beam by gas molecules in its path SEMs must operate at high vacuum. Maintaining the vacuum requires that samples be clean, dry, and stable in a vacuum environment. Second, because the beam electrons deposit charge in the sample, the sample must be conductive. Samples that are not inherently conductive are often made so by depositing a thin conductive coating over the surface. While this solves the conductivity problem it may obscure other information.
ESEMs relax many of the constraints of conventional SEMs. By shortening the path over which the beam is exposed to the sample vacuum environment, they permit the use of partial vacuum. They can be used to observe wet, dirty, outgassing samples while preserving the high-resolution capabilities of SEM. Equally important, they relieve the requirement for sample conductivity. Some of the gas molecules in the beam path are ionized and provide a mechanism for charge neutralization at the sample. Relieved of the requirements for static samples with fixed coatings, ESEMs are also capable of observing dynamic processes and sample interactions, for example, the dissolving of a drug particle in liquid water, or the agglomeration of particles as a function of changes in temperature and humidity.
ESEM Particle Analysis
Figures 1-4 illustrate the application of ESEM to pharmaceutical particle analysis. The sample was an active compound for which the investigators were having difficulty interpreting the light scattering data based on the known dissolution behavior of the material. The particles were not typically spherical. Figure 1 shows a high-resolution image of a representative particle. Figure 2 shows a wide field image used to collect data from a large number of particles dispersed on a filter. Figure 3 is an example of the detailed information reported for each detected particle. Finally, figure 4 compares the particle size distributions derived by the ESEM and light scattering.
[FIGURES 1-4 OMITTED]
In this analysis the sample material is diluted and filtered, and the filters are introduced directly into the ESEM sample chamber. There is no need to dry or coat the filters. Three areas of concern require attention in the development of the analytical process: the selection of the filter medium, the definition of the dilution protocol, and the design of a valid image based sampling strategy.
The filter material is selected to provide optimal discrimination between substrate and particle. Particles are detected on the basis of signal thresholding, i.e., any image pixel with signal strength above a specified level is regarded as part of a particle. The signal used in this case is the Back Scattered Electron (BSE) signal. It consists of beam electrons that are scattered back out of the sample by collisions with sample nuclei. BSE contrast is a function of atomic number, with larger nuclei scattering more electrons and therefore appearing brighter in the image. The challenge is to find a filter medium that has sufficiently different average atomic number from the particles. The optimal material in this case is a commercially available filter (Millipore) with atomic number lower than the drug compound, resulting in bright particles on a dark background.
Sample dilution is designed to provide a maximum number of particles in the field of view without significant particle overlap or agglomeration. The field of view is dependent on the particle size and image resolution, which is, in turn, dependent on the ESEM operating conditions selected for the analysis. Shape determination is assumed to require a minimum of five pixels across the minimum expected particle size.
Sampling strategy for image based counting requires consideration of edge effects. Particles that are contiguous with the edge of an image field are excluded since it is impossible to know that they are completely included in the image. Furthermore, image fields are not adjacent in order to avoid any possibility of overlap and double counting.
Figure 4, comparing the results of ESEM and light scattering, demonstrates the kind of bias that can be introduced by particle shape in light scattering measurements. The particle size distribution is bimodal in both sets of data, however, the larger size mode is smaller in the scattering data than in the ESEM data. The image based ESEM analysis reveals that the particles are generally not spherical and contain sharp edges. The edges tend to scatter light in the same way as smaller particle and so tend to skew the size distribution toward smaller particle sizes.
Light scattering techniques for particle characterization are fast, repeatable and sensitive to both shape and size variations. As such, they will remain the best choice for process monitoring in production. However, light scattering alone cannot provide the detailed qualitative information required to develop new processes or understand excursions in existing processes. These require image-based analysis. Although slower than light scattering, image based analysis has a number of key strengths:
* Direct observation of the particles with the ability to revisit stored data and images for further examination and interpretation.
* Detailed information on particle size and shape.
* Wide dynamic range, capable of analyzing particles from the nanometer to the millimeter scale.
The trend toward smaller particle sizes has pushed image-based analysis beyond the capabilities of optical microscopes. ESEM analysis provides the resolution required for new nanoscale particles without the difficulties and constraints of conventional SEM.
--By: Dr. Arjen Tinke Janssen Pharmaceutica and Ben Lich FEI Electron Optics
|Printer friendly Cite/link Email Feedback|
|Author:||Tinke, Arjen; Lich, Ben|
|Date:||Apr 1, 2005|
|Previous Article:||Apotex Inc.--expansion of the Etobicoke, Ontario facility.|
|Next Article:||Hopper features flared side rails and end stops.|