Until recently, Raman spectroscopy was not widely applied in the pharmaceutical industry. However, in the last few years, developments within the pharmaceutical industry, coupled with improvements in Raman instrumentation, have generated renewed interest in the technique. It is already proving to be a valuable tool in the analysis of pharmaceutical polymorphs and now looks destined to find increasing application throughout the pharmaceutical industry. 

After 80 years of maturing in academic laboratories around the world, Raman spectroscopy has finally begun to make a significant impact in the industrial environment, finding particular application in the semiconductor, chemical and synthetic polymer industries. While still not as widely employed in the pharmaceutical industry as its cousins, near- and mid-infrared (IR) absorption spectroscopies, it has recently found a particular niche in the analysis of pharmaceutical polymorphs. Raman spectra have proved highly characteristic of different polymorphic forms of pharmaceutical compounds and the ability to obtain spectra without the need for potentially polymorph changing sample preparation has proved ideal for applications from polymorph screening to patent litigation. This high profile application of Raman spectroscopy has lead to greater acceptance of, and interest in, the technique within the pharmaceutical industry. With the advent of the US Food Drug Administration's (FDA) process analytical technology (PAT) initiative, Raman spectroscopy is increasingly being considered for applications in the pharmaceutical manufacturing environment. 

So that a reader unfamiliar with Raman spectroscopy can begin to determine its suitability to their own particular application, the basic principles and features of the technique are outlined with particular reference to practical application in a pharmaceutical environment. Comparisons are made with IR spectroscopy with which the reader may be more familiar. Some example applications are presented including the detection and identification of pharmaceutical polymorphs. The potential of the technique as a PAT tool will be discussed separately in Part II. 

A large body of excellent literature exists documenting the basic theory and practical application of Raman spectroscopy, some of which are recommended.

 The essential features of the technique are, however, outlined here for the benefit of those readers unfamiliar with the technique. 			

Raman spectroscopy is a form of vibrational spectroscopy and as such is related to near- and mid-IR absorption spectroscopy.

 Similar to the IR absorption techniques, Raman spectroscopy measures the vibrational frequencies of various parts of a molecule. These frequencies depend on both the bond strength and mass of the bound atoms as well as other factors such as intermolecular interactions. The 'pattern' of vibrational frequencies from a molecule is, therefore, highly characteristic of a given molecular species and, for solid samples, of the crystalline arrangement of those molecules. Raman spectra may be readily recorded from gasses, liquids and solids.

 Figure 1 Schematic overview of the basic experimental arrangement to perform Raman spectroscopy.			

While IR spectroscopy is essentially based on illuminating the sample with a broad range of wavelengths of IR light and measuring which are absorbed, a Raman spectrum is obtained by illuminating the sample with a single wavelength of light from a laser and collecting and analysing the resulting scattered light (Figure 1). Briefly, the basic experimental arrangement to perform Raman spectroscopy includes the following steps: 					

To obtain a Raman spectrum the sample is illuminated with light from a laser.

The molecules in the sample 'scatter' the incident laser light. Some of this scattered light contains detailed information about the molecular properties of the sample.

The scattered light is collected and analysed using a Raman spectrometer to produce a Raman spectrum.

Raman spectra are generally clear, well resolved and rich in features facilitating detailed and unambiguous interpretation.

The spectra contain large amounts of information about chemical composition, intra- and intermolecular phenomena, and the longer range structure of a sample.

The wavelength of the exciting light can lie anywhere from the ultraviolet (UV) to the near-IR ([NIR] with the choice of wavelength used being determined by the details of the application). When the scattered light is collected and spectrally analysed it is observed that a relatively large amount of light is scattered at the wavelength of the incident light; this is known as Rayleigh scattered light and in practice is filtered out and discarded. In addition to this, much smaller amounts of scattered light are observed at various wavelengths shifted with respect to the exciting wavelength; this is the Raman scattered light. Closer inspection reveals that the wavelength difference between the exciting light and each of the various components of the scattered light, when converted to units of frequency, corresponds to the frequency of the molecular vibration giving rise to the scattering. 

Another key difference between Raman spectroscopy and its IR cousins is that although all the techniques measure the vibrational frequencies of a molecule, the fundamental physical process giving rise to the effects is different. In practice this means that while some molecular vibrations may be observed in both the IR and the Raman spectra, often vibrations observed using one technique will be extremely weak or totally absent in  the other. This complimentary nature can have practical consequences. For example, the IR absorption spectrum of water has very strong features corresponding to the O-H bending vibration (frequency ~1655 cm

) and the O-H stretch vibration (frequency 3500 cm

), which tend to obscure the features from other compounds that may be present in solution or in a slurry. By contrast, the Raman signal from water is very weak making the technique particularly suitable for measurements from aqueous samples. 

An important feature of the Raman effect is its low efficiency and the consequent very low intensity of scattered light; for every 10

 photons incident on the sample, typically only one Raman scattered photon is produced. This has a range of practical implications ranging from problems as a result of fluorescent interference from the sample to the requirement for sensitive, low noise detectors. 

The features in the Raman spectrum corresponding to the various vibrational modes of a molecule are generally sharp, well resolved and, for many active pharmaceutical compounds, numerous. This is well demonstrated in Figure 2, which compares the spectra of a number of common pharmaceutical compounds. In this example, it is worth noting the large number of sharp, well-resolved peaks and the many differences between the spectra. Spectra were obtained using a Renishaw Raman microscope with 785 nm excitation and 10 s exposure time. A zoomed in view of the 400–1000 cm

 region of the spectrum is shown. Example characteristic features for each of the spectra are highlighted.

Figure 2 Example Raman spectra from some common pharmaceutical compounds.			

. The instrumentation required to perform Raman spectroscopy is, in principal, relatively straightforward. The instrumental details will depend entirely on the functional specification of the instrument, but typical components are described. 

 The sample is illuminated with a laser of a suitable power, wavelength and stability. Examples commonly employed in modern instrumentation include diode pumped solid state lasers (most often 1064 nm and 532 nm, but many other wavelengths are available) and stabilized NIR diode lasers (typically ~780–820 nm) operating at powers of a few tens of mW to ~1 W. These lasers are generally compact, reasonably robust and reliable. 

. The exciting laser light may be delivered to the sample either via a conventional optical system of mirrors and lenses, or via fibre optic cables, which may extend many tens of meters (note that this is not possible with mid-IR absorption spectroscopy because of absorption in the optical fibre). A lens system focusses the exciting light onto or into the sample and collects the resulting Raman scattered light. It is common to couple lab-based Raman spectrometers to optical microscopes so that delivery and collection of light are performed by a high numerical aperture objective. In this arrangement a micrometer sized sampling volume may be achieved, which is particularly suitable for obtaining spectra from small samples, samples embedded in a matrix or to produce high spatial resolution 'Raman maps' such as that shown in Figure 3. Here, the tablet was sectioned using a microtome to reveal the core. An array of Raman spectra was recorded at 13 μm  intervals from a 1.5 mm

area on the sectioned surface of the tablet. The sampling volume for each spectrum was ~5 mm

and the exposure time for each spectrum was 5 s. Some examples of the recorded Raman spectra are shown. The instrument software was then used to look for characteristic features in the recorded spectra and thus, determine the identity of the material at each point. This information was then used to generate the colour coded map illustrating the distribution of each of the tablet ingredients. It is worth noting that much smaller sampling volumes are possible (by an order of magnitude) with Raman spectroscopy than with IR spectroscopy as a result of the shorter excitation wavelengths used (giving a smaller diffraction limited spot) and because of the fact that all the collimated light emitted by a laser radiation can be readily focussed to an intense diffraction limited spot. This is not the case with traditional, nonsynchrotron IR sources.

Figure 3 Example Raman map of tablet core showing the distribution of APIs and the principal excipient.			

Collecting and filtering the scattered light.

 Efficient collection of the scattered light is essential in Raman spectroscopy because of the weakness of the effect. The collection optics must, therefore, be positioned as close to the sample as possible and have as large an aperture as possible to give a large collection angle (i.e., large numerical aperture collection optics should be used). The collected scattered light must then be spectrally filtered to remove the Rayleigh scattered component otherwise this would swamp the Raman signal. Various types of optical narrow band rejection or edge filters are available to perform this function. 

 The collected Raman light must then be spectrally analysed. This may be done using a low stray light, dispersive spectrometer, with a bandpass filter arrangement or a Fourier transform (FT) type spectrometer. A dispersive spectrometer is generally preferred for excitation wavelengths from the UV to ~820 nm while an FT system is generally required for excitation wavelengths at longer than ~820 nm (e.g., 1064 nm Nd:YAG).

 A sensitive, low noise detector is required to detect the Raman scattered light. For UV-NIR excitation slow scan, cooled charged couple device (CCD) cameras are the detector of choice. Such a detector will, however, typically represent the single most costly component of a Raman system. For some applications it may be possible to employ a point detector or a simple array detector which, while providing only very limited spectral information, will be much less expensive. Array detectors suitable for Raman spectroscopy with IR excitation wavelengths have until recently been prohibitively expensive and so a single element detector coupled to an FT spectrometer is generally used.

These, however, are less sensitive than a system comprising a good dispersive spectrometer and a cooled CCD detector. Recent developments in indium phosphide/indium gallium arsenide phosphide array detectors, however, show promise for IR excited Raman spectroscopy.

. The way in which the various components are combined to build a Raman spectroscopic system is critical to the ultimate performance of the instrument. Careful design is required to ensure system reliability, stability and robustness — parameters of vital importance for operation in an industrial environment. 

. The details of the instrument control and data analysis software will be application dependent. Various proprietary software packages are available that provide a basic interface and often quite sophisticated data analysis facilities coupled with facilities to control instrumentation and tailor the software to an individual application. Increasingly, Raman instrument manufacturers are developing software tailored to the needs of the pharmaceutical industry. 

. Various other features may be required of a Raman system when used in specific applications. For example, when operating in a regulatory environment, automated system performance validation to suitable standards and handling of the data (e.g., compliance with 21 CFR Part 11) are desirable. Instrument robustness, stability and good technical support provided by the manufacturer are important for 'mission critical' applications. 

. There are a number of established manufacturers of Raman instrumentation.

 These offer a range of different instruments targeted at different applications including Raman microscopes and instruments designed for industrial and manufacturing applications. Alternatively, a number of companies exist that will design and build custom instrumentation optimized for a particular application. 

 A given solid state pharmaceutical compound can often have different polymorphic forms. This can affect the ease of manufacture, efficacy and stability of the final formulation. Detection and identification of different polymorphic forms is, therefore, important throughout the drug development and manufacturing process, for example: 					

During the drug discovery/development process so that the optimum form for eventual manufacture can be selected. Analytical identification of this form will be an integral part in its patent protection. 

During scale up and manufacture it is important to monitor the material being produced to ensure the drug is of the correct polymorphic form for inclusion in the final product.

Once at market, it may be desirable to determine the form of a competitor's material to help establish whether any patent infringement has occurred. 

The molecules in different polymorphs experience different intermolecular forces resulting in modifications of both the frequency and intensity of some of the modes of vibration observed in the Raman spectrum. In practice, Raman spectra have proven to be very sensitive to the polymorphic form of a material with often substantial spectral differences evident between polymorphs. The sharp, well resolved features typical of pharmaceutical compound spectra also facilitate the identification of more subtle spectral differences which may, for example, indicate the presence of low levels of one polymorph within another. 

In the Ranitidine spectra shown in Figure 4 there are more than 40 clearly resolved peaks in the limited spectral range shown. A zoomed in view of the 900–1400 cm

 region is shown and example characteristic features highlighted for each of the polymorphs.

Figure 4 Example Raman spectra from two different polymorphs of Ranitidine HCl obtained from tableted formulations from different manufacturers.			

Note the large number of sharp, well-resolved peaks and the large number of differences between the two forms demonstrating the ability of Raman spectroscopy to unambiguously discriminate between and identify different polymorphs of the same material. The highlighted characteristic features for each polymorph do not overlap and would thus be potentially well suited to use in a univariate calibration model for the quantitative analysis of mixed samples. 

The spectra were recorded with a Renishaw Raman microscope with 785 nm excitation and 15 s exposure.  The sensitivity of Raman spectroscopy to polymorphic form has been demonstrated for a number of pharmaceutical compounds.

The absence of the need for sample preparation and short spectral acquisition times makes Raman spectroscopy particularly suitable for high throughput identity verification of raw materials. The facility to record spectra through optically transparent containers/windows gives the potential to test materials without opening their containers or in-line as they are fed into the production line. 

. The high spatial resolution (small sampling volume) available when a Raman spectrometer is coupled to an optical microscope enables chemically specific spectral maps of tablet cross sections to be generated.

 These can typically have a spatial resolution of ~0.5–2 μm and the highly characteristic, well resolved nature of Raman spectra enables the simultaneous mapping of several ingredients. An example Raman map of a commercially available analgesic formulation is shown in Figure 3. The spatial distribution of the three active pharmaceutical ingredients and the principal excipient is clearly revealed. These measurements are straightforward to perform and can give information about the homogeneity of mixing on a micron length scale. 

Contaminant identification and cleaning verification.

 The high spatial resolution of a Raman microscope makes it particularly useful for identifying small particulates such as contamination within formulated products or on swabs as part of cleaning verification. 

Raman spectra are rich in information about the molecular nature of a sample. In particular the Raman spectra of crystalline pharmaceutical compounds typically contain many well-resolved features and are highly characteristic of a given material. These spectra have also been shown to be highly sensitive to the different polymorphic forms of pharmaceutical compounds. Little, if any, sample preparation is required to record a Raman spectrum making it particularly suitable for polymorph identification or high throughput raw material testing. Raman spectra may be recorded from very small sample volumes enabling micron-sized particles to be identified or micron spatial resolution chemically specific maps to be generated. 			

Part II of this article will focus on the potential for Raman spectroscopy as a PAT tool. The technique's strengths and weaknesses are discussed regarding its implementation within a pharmaceutical manufacturing environment. 

 is a senior scientist specializing in spectroscopic characterization of pharmaceuticals and biologics and 

 is a senior scientist specializing in spectroscopic and optical instrumentation both at Avacta Ltd, UK.

Modern Raman Spectroscopy: A Practical Approach

 (John Wiley and Sons Ltd, Chichester, UK, 2004). 			

Analytical Applications of Raman Spectroscopy

 (Blackwell Science Inc., Malden, MA, USA, 1999). 

3. I.R. Lewis and H.G.M. Edwards (Editors), 

Handbook of Raman Spectroscopy: From the Research Laboratory to the Process Line

 (Marcel Dekker Inc., New York, NY, USA, 2001). 

Articles

Features

Detection and identification of different polymorphic forms is, therefore, important throughout the drug development and manufacturing process.

Raman Spectroscopy for Pharma. Part 1: Principles and Applications

Novel Peptide Vector Carries Drug to Human Brain

Re-engineered drugs create new treatments

 A new technology from Synt:em has demonstrated, for the first time, its ability to transport a drug across the human blood brain barrier (BBB). The method, which employees Pep:trans, a range of peptide-derived vectors to which compounds can be conjugated, was used to develop Syn1001. The drug, formed by chemically attaching active morphine metabolites to the vector, penetrated the brain and relieved pain, proving the Pep:trans technology concept — that drugs could be re-engineered to form alternatives that cross the BBB.			

				"We have now opened up a way to rapidly develop new drugs that address the huge market of central nervous system diseases. We believe that Syn1001 is well positioned as a potential product to treat postoperative pain," said Michel Kaczorek, founder and CEO of Synt:em.  				

 						To speed-up time between research outcomes and development of genome-based drugs, PharmaLogicals Research Pte — a collaboration between Chungai, Mitsui and CIEA — has formed a research system that integrates drug design strategy and pathological studies with cutting-edge advances in genomic research. Based at the University of Tokyo's Research Centre (RCAST), the Forerunner Pharma Research Co Ltd, as it will be known, is anticipated to make use of the available research in pathological proteomics from PharmaLogicals and also further develop its collaborate research with RCAST and other institutes. 						

			Arakis and Vectura, who joined forces to develop the long acting antimuscarinic agent AD 237 for the treatment of chronic obstructive pulmonary disease, have added Novartis to their team. Novartis will be responsible for developing the drug, presently in Phase II clinical trials, as a monotherapy as well as in combination with its long acting beta2 agonist, QAÂ±49, also in Phase II development.  

 Using ultrasound, Japan-based Kagoshima University and KSL, claim to have developed a new transdermal drug delivery system that could make injections a thing of the past. 			

The new method should allow a wider range of drugs, such as the larger-sized insulin, to pass the skin barrier without the need for injection. The technique, which the companies hope to have on the market in the next five years, has already been used to pass a fever-reducing drug across the skin of mice, proving its efficacy. 

A new technology from Synt:em has demonstrated, for the first time, its ability to transport a drug across the human blood brain barrier (BBB).

R & D News

Pharmaceutical manufacturing processes involve a variety of chemical or biological materials. The quality control (QC) of materials and formulations for pharmaceutical use include, for example, the confirmation of a chemical purity, polymorphism and water content. Other important characteristics are physical and functionality-related properties, which can vary between different grades of the same chemical entity. Such characteristics include dissolution rate; physical state (dispersed/aggregates); thermal stability (thermal transition); and particle size. This is true, in particular, for biologically active macromolecules such as proteins, polysaccharides or synthesized polymers. The specific physical and functionality-related properties can be tested by different methods such as ultraviolet (UV)/visible spectroscopy, nuclear magnetic resonance (NMR), laser diffraction, calorimetry and dynamic rheology. All have their limitations and, therefore, the search for new, fast and effective methods for wide-range characterization of material (raw materials, intermediates, excipients and active ingredients) is a hot issue in drug production quality. 

High-resolution ultrasonic spectroscopy (HR-US) is a nondestructive technique with enormous potential for the analysis of materials and formulations used in the pharmaceutical industry. This technique is based on precision measurements of parameters of acoustical waves at high frequencies propagating through materials. Unlike current analytical methods, optical transparency is not required because ultrasonic waves propagate through opaque samples. These instruments require small sample volumes, down to 0.03 mL, and give excellent resolution. They can be used for analysing composition; aggregation; gelation; micelle formation; crystallization; dissolution; sedimentation; enzyme activity; conformational transitions in polymers; ligand binding; antigen-antibody interactions; and many other processes that play a key role for functional properties of material used in drug production. Capable of dealing with a wide range of samples and dynamic processes, HR-US generates product quality information in real process time. It allows fast analysis of formulation consistency; batch-to-batch variation, stability assessment and so forth. 			

Two independent parameters, ultrasonic attenuation and ultrasonic velocity, are measured in ultrasonic spectroscopy. Ultrasonic attenuation is determined by the energy losses in ultrasonic waves and can be expressed in terms of the high-frequency viscosity of the medium or its longitudinal loss modulus. This allows the analysis of kinetics of fast chemical reactions and the microstructure of materials including particle sizing, aggregation, gelation, crystallization, and other processes and characteristics. Ultrasonic velocity is determined by the density and elastic response of the sample to the oscillating pressure in the ultrasonic wave and thus, can be expressed in terms of compressibility or storage modulus. This parameter is extremely sensitive to the molecular organization, composition and intermolecular interactions in the analysed medium, and is responsible for the majority of applications of HR-US for analysis of chemical properties of materials. HR-US spectrometers, with their superior resolution, small sample volume and robust design, are the first commercial instruments that allow the user to enjoy the full potential of high-resolution ultrasonic analysis. Standard sample cells for the HR-US spectrometers have a 1 mL capacity. A range of cells is also available, which include: 					

flow-through cells that allow analysis in-flow

cells with small volume capacities (as low as 0.03 mL).

All measurements can be made at temperatures as low as –60 °C and as high as 120 °C.  

The paper describes some applications of HR-US for different types of material analysis in pharmaceutical production. Several examples are illustrated including the measurements of the particle size in drug emulsions with different polymer coating; monitoring the dissolution of drug substances; the characterization of thermal stability of polymers used in drug formulations; and the analysis of water content in different forms of the same drug compound. 

Control of particle size in the dispersed phase is a critical issue in pharmaceutical formulations. Particle characteristics can affect many different areas including inhalation delivery systems, tablet dissolution characteristics, formulation quality, and solubility or absorption. Batch-to-batch variations in particle size can lead to unpredictable variations in the life span and stability (shelf life/heat stability) of a product. Traditionally, the characterization of the particles in dispersion is made by optical methods such as light scattering. This means that the sample must be diluted to reach optical transparency and avoid multiple scattering effects. However, the application of HR-US measurements allows direct analysis of the particle size and their volume fraction in the dispersions (even concentrated samples, e.g., 40%), thus avoiding the problems associated with dilution (such as change in the particle size and aggregation rate). 							

In the current example, HR-US was used to analyse the effect of different polymer coatings on the sedimentation rate and the evolution of the particle size in drug suspension. The samples of drug suspensions were prepared using hydroxypropyl cellulose as a drug carrier system. The drug particles in Sample 1 were coated with polymer, which had a higher molecular weight (approximately twice) compared with the polymer used in the preparation of Sample 2. Each of the samples were shaken vigorously before the measurements and loaded into an ultrasonic cell. The sedimentation processes were monitored continuously over 30 min.  

Sedimentation is reflected in ultrasonic measurements as a change in velocity and attenuation with time. HR-US software deconvolutes these changes and provides the change in particle size and volume fraction during the sedimentation process. Evolution of particles' size and their volume fraction in two types of drug suspension is compared in Figure 1. Initial particle size (after loading the sample into the cell) in Sample 1 was 0.8 μm, which is 40% larger than the size in Sample 2. However, the particle size in Sample 2 increases faster with time as a result of the aggregation. Particle growth is accompanied by a decrease in volume fraction, which can be attributed to the sedimentation, by about 0.05% in Sample 1 and 0.3% in Sample 2 within 25 min. Both processes (sedimentation and aggregation) are more evident in Sample 2, indicating lower stability compared with Sample 1. One of the possible factors for the higher stability of Sample 1 can be a 'thicker' polymer layer on the drug particles in this sample, which protects the particles against aggregation. 

 Figure1. Evolution of changes in volume fraction of drug particles and particlesÃÂ´; size calculated from the ultrasonic velocity and attenuation data measured in two undiluted drug suspensions with different polymer coating at 25 Â°C. 			

The effect of dilution on the evolution of the drug particles' size was analysed by measuring the sedimentation profile (similar to those in Figure 1) in diluted suspensions. Figure 2 shows that average particle size decreases with dilution. This decrease can be explained by the effect of the particle concentration on aggregation. Stirring the suspension before the measurements are taken results in the breakage of the particle aggregates into smaller particles. In undisrupted suspension, the particles' growth in size is a result of an aggregation. However, because the aggregation rate decreases at lower drug concentration, the average particle size measured after the same time interval (i.e., 25 min) is smaller in undiluted suspension. Figure 2 also illustrates the effect of dilution on the particle size depends on the particle coating. In very diluted samples (e.g., 1:20) the drug particles are larger in Sample 1. This can be attributed to a 'thicker' layer formed by a higher molecular weight polymer on the original particles. The results show that an increase in the drug concentration overall enhances the aggregation and that this process is less pronounced in Sample 1 because of a more effective protection polymer layer on drug particles as previously discussed. Optimization of drug coating and the improvement of sample stability can be effectively controlled using high-resolution ultrasonic measurements. 

Figure2. Change in size of the drug particles with concentration for two different polymer coatings as determined by the HR-US measurements. The particle size was measured after 25 min of the sedimentation and giving as relative to one in undiluted sample 1.			

Control of water content in raw material and active compounds is one of the key issues in drug production. HR-US spectrometers allow analysis of water concentration using high-precision velocity measurements. The ultrasonic analysis is based on the fact that for many solutions and mixtures, the ultrasonic velocity is proportional to the concentration of solute component. When the contributions of the components to the ultrasonic velocity are known (e.g., after calibration) the concentration of the components can be precisely determined from the measured ultrasonic velocity. This approach was applied to measure water content in the crystalline and amorphous forms of the drug active compound.  			

Two types of drug compounds — amorphous and crystalline — were dissolved in methanol at 0.13 g/cm

 concentration and loaded into ultrasonic cells of the HR-US 102 spectrometer (Ultrasonic Scientific Ltd, Ireland) for the measurements of relative ultrasonic velocity (using methanol as a reference). The comparison of ultrasonic parameters of two drugs solutions (at the same nominal concentration of drug powder, mg/cm

) at 20 °C is given in Figure 3(a). The data demonstrates that samples can be easily distinguished by their ultrasonic profiles. The measured difference in ultrasonic velocity is several orders higher than the accuracy of the HR-US 102 spectrometer. Water content was calculated from the measured ultrasonic velocities and the calibration curve, and is shown in Figure 3(b). The total amount of water in the samples is between 1–1.5%. Water content is higher in the amorphous form of drug compounds compared with the crystalline version by approximately 0.5%. 

Figure 3. Water content in two forms of the same drug substance measured with an HR-US 102 spectrometer. The values of the ultrasonic velocity are given for 1 mg/mL drug suspension relative to methanol.			

An example of the practical use of HR-US is monitoring thermal transitions in polymers used in different drug formulations. Hydroxypropylmethylcellulose (HPMC) is used to encapsulate hard capsule products. This polymer dissolves slowly in cold water to form a viscous colloidal solution and upon heating becomes insoluble in water forming a thermally reversible gel. This use is petitioned as an alternative to gelatin (animal based) capsules. HPMC has many other uses as an emulsifier, thickening agent, stabilizer, gellant and suspending agent.  			

In this study, the HR-US 102 spectrometer was used for comparative analysis of the gelling transition in two samples of HPMC (HPMC 1 and HPMC 2) that differ in the hydroxypropyl content. The solutions of HPMC samples in water (2 w/v) were loaded into an ultrasonic cell and their thermal gelling profile was measured from 40–90°C at a heating rate of 0.5°C/min.

Figure 4. Monitoring of thermal transition in two types of hydroxypropylmethylcellulose. Evolution of ultrasonic velocity in polymer solution upon heating from 40-95 Â°C: (a) original data and (b) after the subtraction of baseline.			

Figure 4 shows the change in ultrasonic velocity with temperature in solutions of two different HPMC samples. The velocity profile at low temperature exhibits a linear decrease with increasing temperature and was used as a baseline. Subtracting a temperature dependent baseline shows the details of the gelling transitions in two samples. The decrease in velocity in this gel transition is caused by a decrease in the hydration level of the atomic groups of the HPMC molecules (the compressibility of water in the hydration shells is typically less than the compressibility of the bulk water, and, therefore, the transfer of hydration water into the bulk water reduces the ultrasonic velocity). The gelation temperature can be determined as the point of inflection on the transition curves in Figure 2. The data show that the gelation temperature in HPMC 1 (56°C) is lower than in HPMC 2 (72°C). The overall magnitude of change in ultrasonic velocity characterizes the degree of cross-linking and is an indicator of the gel strength. According to the ultrasonic data, HPMC 1 forms the strongest gel networks in comparison with the HPMC 2 sample.  

Thus, HR-US measurements allow analysis of the gelling profiles, detection of the gelling temperatures and characterization of cross-links in the gel.  

Dissolution rates of raw materials, active ingredients and excipients are important functional characteristics that determine drug efficiency. HR-US offers a direct and simple method for the analysis of the dissolution rate. In the following example, the dissolution rate of two different tablets in an acidic buffer, mimicking the gastric acid environment, was monitored using the HR-US 102 spectrometer.  			

Figure 5 presents the dissolution rate of two tablets determined from the ultrasonic velocity data. The gradual dissolving of solid drug components results in the growth of solute concentration, thus changing the ultrasonic velocity. The drugs dissolved completely after 2 h. It can be seen from Figure 5 that Tablet 1 initially dissolves at a much slower rate than Tablet 2. Half of Tablet 1 is dissolved in the first 12 min, whereas it takes about 35 min to dissolve 50% of Tablet 2. However, the dissolution rate of both tablets is found to be similar after 20 min. One of the advantages of the ultrasonic method for measuring the dissolution rate is that it allows measurements of practically any compound in various environments (e.g., different temperatures, pH and solvents).

Figure 5. The rate of dissolution of two tablets as monitored with high-resolution ultrasonic velocity measurements.			

The examples described show the high potential of HR-US for analysing a broad range of important properties of different raw materials, excipients and final formulations used in drug production. The arrival of the next generation of ultrasonic spectrometers and analysers will have a profound impact on the efficiency and profitability of pharmaceutical processes.			

Evgeny Kudryashov, Cormac Smyth, Breda O'Driscoll 

are senior scientist, product manager and managing director at Ultrasonic Scientific, Ireland.

 is head of the laboratory of Physical Chemistry of Biocolloids at the Department of Chemistry, University College Dublin, Ireland.

HR-US is a nondestructive technique with enormous potential for the analysis of materials and formulations used in the pharmaceutical industry.

Analysing Raw Materials and Formulations Using High Resolution Ultrasonic Spectroscopy

Clean rooms are now used in a variety of industries, including microelectronics, nanotechnology, defence, pharmaceuticals and biotechnology. They range from small rooms and large suites to modular purpose-built structures feeding a comprehensive and complex maze of utilities and serviced equipment. 			

The principal function of a clean room is to protect the manufacturing product from contamination. In the pharmaceutical sector the commercial survival of the manufacturer depends on the integrity of the finished product. Therefore, it is important to identify a potential source of contamination, which could include the working environment itself.  

However, it is important to remember that the quality control and standard operating procedure (SOP) requirements of a clean room are prerequisites for a successful clean room operation. A clean room can be defined as an environment where the supply, distribution and filtration of clean air and the material of construction are regulated to meet the appropriate cleanliness levels required and defined by the governing authorities to execute a validatible clean room. 			

Clean rooms within the pharmaceutical industry cannot be considered in isolation. The decision to use designated space within a production facility cannot be taken lightly. They are expensive to build and operate, which gives the design team the task of balancing a number of factors:  					

What level of clean room environment is required?

Are the logistics surrounding the proposed facility affordable? 

The design team will have to consider these options at the conceptual stage because different types of clean room design require different disciplines to lead the design and layout process. 			

Primary pharmaceutical clean room environments. 

For these facilities the process engineering equipment and piping layouts determines the strategy because they are key elements of the manufacturing process (i.e, the process will be the lead discipline). The clean room is likely to be a small offloading, vessel charging or dispensing suite. 

Secondary pharmaceutical clean room environments.

 Here, the architecture of the room layouts and the air environment are the 'manufacturing vessels' where products and people populate a given space. The architecture or mechanical services are the crucial factor in getting a correct synergy in people and product traffic flows 

Biopharmaceutical clean room environments.

 For these facilities, the process vessels and equipment, the room layouts, and the flow of people are equally important to the process. Both processes and architecture are the lead disciplines to produce an integrated solution. 			

Possible cross-contamination issues should be eliminated at the early stage of the project. The project sponsor should ensure that all relevant personnel from the production, quality control, logistics, and maintenance departments, as well as engineering, are involved in the conceptual stages of a design. This avoids reworking design layouts later because the whole company has taken ownership. Therefore, the input from a wider 'team' reduces costly errors. 

This team, which can include the design contractor, develop the initial user requirement specification (URS). This needs to be a short definition of the processes, equipment, operations, capacities and environmental criteria in the clean rooms to be designed. 			

 The architectural room layouts and the heating ventilation and air conditioning (HVAC) environment are fundamental to a contained suite of laboratory clean rooms. Both architectural and HVAC disciplines take the lead to produce an integrated solution. When embarking on a project of this nature it is a prerequisite that the project sponsor has all the relevant departments in place so that the design team have a good understanding of the task ahead. This team will encompass all the ideas from the collaborative URS from which the design can begin. 

In the URS the design team needs to review the flow of material and people around the clean rooms to assess the optimum layout for regulatory compliance (current good manufacturing practice [cGMP]), efficient operation and to minimize cross contamination.  			

It is easier to simplify these flows in new facilities, but difficult in retrofit situations where compromises may be needed to accommodate space or cost constraints. In this situation procedures and controls may need to be established to avoid cross contamination where waste, people, raw materials and finished goods share single corridors.  

In summary the clean room design team must resolve all issues within the scheme design stage. Figure 1 summarizes the clean room design thought process from conception to completion. 

 Figure 1: The clean room design thought process.			

The key goal of layout design is to maximize the technical and lower classification areas around the main clean room suite by locating and specifying process equipment so they can be placed to allow maintenance from the non-clean room area. 			

There are two main benefits to this; first it minimizes the costly clean room partitioning system; and second it minimizes the maintenance procedures and costs. 

This is not always easy to do and will be further hampered by the standard designs of equipment. Many pharmaceutical equipment suppliers are developing new equipment models and it is worth researching these at the start of a project as they may reduce capital and running costs. 

Blow fill seal filling machine technology is an example of this. Up until a few years ago the whole machine had to be inside the clean room. Now suppliers offer machines that have 50% in a technical non-clean room environment and 50% in the clean room.  

The layout should keep personnel and materials separate. If, because of layout constraints, these must be combined then SOPs will have to be put in place to control them. When planning higher classification clean rooms (Class 5 and Class 7), sufficient space must be allowed in the layout for incorporating multiple change steps (different clothing/change regimes that are required when entering a clean room classification that is higher or lower [step up/step down]). In the case of some Class 5 suites, and containment level 3 and 4 suites, separate exit and entrance routes are required to stop cross contamination of clothing. In smaller higher classification suites, changing rooms, airlocks and transfer hatches occupy a high proportion of the layout. Therefore, it is important to include these costs in any project approval budgets at the conceptual stage. 

As a general guide, Table 1 shows which clean room environment the main product groups require. This is only a guide, and a detailed assessment should be made (and defined) with production, quality assurance personnel and regulatory authorities before the clean room design starts, preferably at the front end design (FED) stage.

Table 1: A typical list of the main products groups and their clean room requirements. 			

Given the wide variety of products and systems on the market, the type of product/finish selected can be a reflection of personal preference and/or budget. 			

However, the overriding philosophy must be to provide a clean room fabric that will include the following criteria: 					

A hard impervious smooth finish, with no sharp edges or angles, to prevent particle generation.

Internal coved corners, angles and equipment plinths to make the surface easier to clean.

Table 2 compares the construction range of clean room finishes against associated processes.

Table 2: Comparison across the construction range of clean room finishes against associated processes. 

There are four basic types of clean room construction: 					

solid core panel monoblock and factory applied finish (type 3)

hollow core 'system factory panel system (type 4).

Generally types 3 and 4 offer more flexibility and are constructed quicker. Types 1 and 2 are more economic, but come with time penalties and less flexibility during construction. The type of wall construction must be defined early in a project (FED), as it affects the total project cost, timetable, and engineering and detailing costs. 

Table 3 shows the comparative cost ranges for the three main clean room classifications used in pharmaceutical facilities. They exclude the shell construction, equipment installation and validation costs, which vary between projects. 			

It is advisable to conduct a concept study or front end engineering study to establish accurate costs because they vary considerably with facility specification, room size, layout and location. 

Table 3: Typical comparative cost ranges for the three main clean room classifications used in pharmaceutical facilities. 

A common mistake made by many companies is to assume that the m

 price from a previous project can be used for budgeting purposes for capital approval only to find that it is inadequate. This will then force the design team to design and specify within the budget available. 

Early in the research and design phase of project the procurement route to be taken must be reviewed by the design team. This will affect the level of detail design done by the project and by the chosen clean room installation contractor. This decision can ultimately affect whole project costs. 			

Table 4 reviews the procurement method for engaging either a clean room contractor to take on the design process or a dedicated consultant engineering team, with the advantages and disadvantages of both. The construction community is, however, currently developing subtle variations on these options. 

Table 4: Procurement routes and their considerations. 

				The ultimate role of a clean room is to exclude the external environment; create a class of cleanliness required for the process by dilution or displacement of the contamination generated; house the equipment or process in an envelope that is fit for its intended purpose; and control and manage the flow of personnel and materials through the planning and running of the suite. 							

Possible cross-contamination issues should be eliminated at the early stage of the project. The project sponsor should ensure that all relevant personnel from the production, quality control, logistics, and maintenance departments, as well as engineering, are involved in the conceptual stages of a design.

Designing & Implementing Pharmaceutical Clean Rooms

Recent work has shown that it is possible to prepare pellets containing a high proportion of a self-emulsifying system as a potential method to convert a liquid into a solid dosage form.

When a poorly water soluble drug was formulated in such a system, the product was equally bioavailable when administered to dogs.

 The system reported involved the use of a mixture of equal parts polysorbate 80 USNF (United States National Formulary) and mono- and diglycerides USNF.  

This mixture forms one of many possibilities for a set of self-emulsifying systems, which can be used for the oral drug delivery of solid compounds.3 So was the system studied a specific combination of self-emulsifying system and microcrystalline cellulose (MCC) or there are other self-emulsifying systems that will function in a similar manner? A further combination of self-emulsifying systems has been compared with the original systems to answer this question. Because the inclusion of lactose in the formulation reduced the ability to incorporate the self-emulsifying system and did not enhance pellet preparation, this ingredient was excluded from the design.  			

Polysorbate 80 USNF (Tween 80) (P) and sorbitan mono-oleate (Span 80) (S) were supplied by Honeywill and Stein (UK). The mixture of mono- and diglycerides USNF (Imwitor 742) (M) was supplied by Condea (Germany) the soyabean oil (SB) was supplied by Croda Chemicals Ltd (UK), while the MCC was supplied by FMC Corp (Avicel PH101; Ireland). The water used throughout was prepared by reverse osmosis.  			

From these materials, three self-emulsifying systems were prepared with the following compositions:  					

A mixture of equal parts by weight of mono- and diglycerides (M) and polysorbate 80 (P) to give a composition MP.

A mixture containing 70% w/w mono- and diglycerides(M), 20% w/w polysorbate 80 (P) and 10% sorbitan mono-oleate (S) to give a composition MPS.

A mixture containing 30% w/w soyabean oil (SB), 49% Polysorbate 80 (P) and 21% Sorbitan mono-oleate (S) to give a composition SBPS.

These mixtures were dispersed with 10, 20, 30 and 40% w/w water to provide self-emulsifying systems, the appearance of which is recorded in Table 1. The mixtures were produced by agitating with a magnetic stirrer until an even texture was achieved. 

Table 1 Consistency of self-emulsifying systems mixed with water.  

All the mixtures were used on the same day that they were prepared.  

To prepare pellets, the appropriate quantity of fluid was added to 50 g of MCC in a mortar and thoroughly mixed with a pestle to provide an even consistency. The correct quantity of fluid required was determined by trial and error. In some instances, the quantities of the self-emulsifying system and the water were added separately to the MCC, the water being incorporated prior to the self-emulsifying system. 

The wet powder mass from these mixtures was extruded through a 1 mm diameter die, 6 mm in length attached to a 25 mm diameter barrel of a ram extruder. The ram was attached to the crosshead of a universal testing instrument (MX 50; Lloyd Instruments, UK) and extruded at a cross head speed of 200 mm/min. The force required to cause steady state flow was monitored by the load cell, the output of which was fed to a computer and displayed on a chart recorder.

 The extrudate, about 80 g, was spheronized for 10 min on a 120 mm diameter spheronizer (G.B. Caleva Ltd, UK) fitted with a cross hatch plate rotating at 1880 rpm. The resultant pellets were dried at 40 °C in an oven (Gallenkamp Hot Box, UK). 			

The particle size of the pellets and its distribution was determined using a set of British Standard sieves (a square root 2 progression of sizes), by agitating for 10 min with a sieve shaker (Endecotts Ltd, UK). The mechanical strength of the pellets was determined by a diametral compression test with a strength tester (CT5; Engineering Systems, UK) at a platen speed of 1 mm/min. The values are the mean of 10 pellets taken at random from the relevant size fraction. The shape of the pellets was assessed with an image analyser (Seescan Ltd, UK) with a black and white camera (CCD-4, miniature video camera, Rengo Co. Japan) connected to a zoom lens (18-108/2.5 from Olympus, Germany) by the method described by Podczeck and Newton.

Table 2 Proportions and properties of pellets produced with mixtures of a self-emulsifying system containing MP and water and MCC.							

A minimum of 100 pellets selected at random from the appropriate size fraction was assessed. The disintegration time of the pellets was measured with British Pharmacopoeia (BP) disintegration apparatus (Copley, UK) fitted with 1.0 mm aperture mesh. The tabled values are the mean of six pellets tested individually.  

Table 3 Proportions and properties of pellets produced with mixtures of a self-emulsifying system containing MPS and water and MCC.			

 none of the self-emulsifying systems, when added to MCC without the presence of water, produced a wet powder mass that could be extruded. As pressure was applied to the mass with the ram, fluid was expressed from the die and the force increased to excessive levels. As soon as water was present at a 10% level, the mass could be extruded in all the three systems tested. The results for the steady-state extrusion force, particle size and size distribution characteristics, shape, mechanical strength and disintegration, are presented in Tables 2–4.  

Table 4 Proportions and properties of pellets produced with mixtures of a self-emulsifying system containing SBPS and water and MCC.						

It is evident from these results that it is possible to prepare pellets from all the compositions tested, thereby extending the generality of the previous studies. There are slight differences in the quantity of fluid, which was required to produce pellets by the different self-emulsifying systems, but these were of the same order of magnitude, ranging from 70–80% of the quantity of MCC. In all instances, the pellets tend to be larger than those that would be produced using a totally aqueous system with the same diameter and length die, and for those made at the higher level of fluid, the modal size exceeded 2.0 mm. Generally the modal size was the 1.4–2.0 mm size fraction and it was usually possible to produce pellets in which an excess of 70% was in this size fraction. It was more difficult to produce the smaller size of pellets when the systems contained 90% of the self-emulsifying systems.  

In terms of shape, only on two occasions did the value of the shape factor e

 fall below 0.55, a value that indicates that the pellets are approximately spherical.

 These two sets of pellets were formed with the 90% level of self-emulsifying systems and the highest quantity of these used. Reducing the quantity of fluid used in the formulations could produce better pellets. Many of the pellet formulations had values of e

 greater than 0.60, a value which cannot always be achieved with aqueous systems. 

This high quality of pellet roundness is surprising considering the high extrusion forces required to extrude the wet masses. Equivalent aqueous formulations using a die of this length and diameter would not be expected to produce acceptable pellets at forces above 10 kN. Such extrudates would produce pellets that are elongated rather than round. Thus, more than a simple measure of the extrusion force is required to evaluate the consistency of wet powder masses used in extrusion/spheronization.  

Another intriguing phenomenon is the ability of the systems to function, both when the water is added directly to the MCC or to the self-emulsifying system. The extrusion force and the properties of the pellets produced for such systems are not different, whatever self-emulsifying system was used (Tables 2–4).  

Another property that could be important if a drug was incorporated into the pellets is whether or not they disintegrate. When tested by a BP procedure, pellets from all the formulations and conditions disintegrated within 15 min.  

Generally the higher the water level of the formulation, the longer the disintegration time. Those pellets produced from the self-emulsifying systems containing soyabean oil, Tween 80 and Span 80 (SBPS) had the longest disintegration times while those prepared from the mixture of mono- and diglycerides with Tween 80 (MP) had the shortest. 

As disintegration times increase with an increase in water used to produce the product so does the mechanical strength for all these systems (Tables 2–4). There is little difference between the strength of pellets prepared from the three different component self-emulsifying systems. In all cases, the pellets prepared with the lowest level of water in the fluid are the softest. These pellets not only have the lowest breaking load they tend to be the most deformable. Hence these could be the most difficult pellets to handle if they are to be filled into hard gelatin capsules. 

In general, it is preferable to have the highest level of self-emulsifying component in the pellet to obtain the potential for drug incorporation. The results in Tables 2–4 show a reduction in achievable self-emulsifier system content from 40–30% when the water content is increased from 10–40% in the wet mass. By use of either 20 or 30% water, the levels of self-emulsifying systems can be maintained in the region of 35% within pellets, which disintegrate within 10 min. There is obviously a compromise to be made in terms of water incorporation to maximize emulsifier content, strength and minimize disintegration times.  

The preparation of pellets by extrusion/spheronization, which can contain up to 40% of self-emulsifying systems and are of an appropriate pharmaceutical quality has been demonstrated for other self-emulsifying systems. It has been confirmed that water must always be present to enable pellets to be manufactured. A consideration of mechanical properties and self-emulsifying content must be undertaken, but levels in the region of 35% of self-emulsifying content of the pellet weight are achievable, allowing the potential administration of poorly soluble drugs as solid dosage forms.  					

is the emeritus professor of pharmaceuticals and 

 is a post-graduate research student both at University of London, UK. 

 is a senior director both at Merck Sharp & Dohme, UK.

4. P.J. Harrison, J.M. Newton and R.E. Rowe, 

This high quality of pellet roundness is surprising in view of the high extrusion forces required to extrude the wet masses.

Pharmaceutical Technology Europe-06-01-2005

Raman Spectroscopy for Pharma. Part 1: Principles and Applications

R & D News

Analysing Raw Materials and Formulations Using High Resolution Ultrasonic Spectroscopy

Designing & Implementing Pharmaceutical Clean Rooms

Formulation Variables on Pellets Containing Self-Emulsifying Systems