Manufacturing Benefits of Hyperspectral and Raman Imaging Techniques

Manufacturing

Applications of hyperspectral and Raman imaging techniques have now By David Bannon at Headwall Photonics been extended to pharmaceutical production where they can be used to increase production yields and provide companies with a competitive manufacturing advantage.

Since their development over 25 years ago, spectral be natural, like sunlight, or man-made – for example, imaging techniques have gradually expanded into more a monochromatic beam of light from a UV/IR lamp or commercial application areas such as pharmaceuticals. a high-energy laser. The source of radiation is directed Recent breakthroughs in the design and performance onto the surface of a sample and – by utilising a of optical components have solved many of the series of collecting, dispersing focusing and detecting limitations associated with older hyperspectral and optical components – the unique spectral information Raman imaging systems. As a result, there is now from the sample can be captured and then used to significant potential for in-line or at-line applications identify the individual components of a material. where imaging performance, reliability and high-speed This technique is called multi- or hyper-spectral data transfer are the defining system parameters (1). The imaging (HSI), because it typically involves collecting US FDA’s Process Analytical Technology (PAT) initiative multiple spectral channels and spatial bands in the readily lends itself to the benefits of spectral imaging electromagnetic spectrum, anywhere from 200nm techniques, as does the need to implement a ‘quality by to 12µm. design’ infrastructure. When used to scan, for example, a wide conveyor of By developing spectral imaging equipment that covers moving tablets or pills, multiple two-dimensional the entire wavelength range – from ultra violet (UV), (spectral and spatial) image frames are captured through near infrared (NIR), up to the long wave infrared as tablets pass by the hyperspectral imager; these (LWIR) – sensor manufacturers can address commercial individual frames are taken at very high speed and demand for imaging in harsh, critical application areas. are stacked like a deck of cards to produce a data The commercial growth of these imaging techniques file commonly called a hyperspectral data cube. The has exploded in the last five years, mainly in their value of each pixel within this hyperspectral data cube ability to solve real-world problems associated with, represents the spectral intensity of that pixel’s small for example, pharmaceutical quality, forensic science, field of view on the scene. human tissue analysis, non-invasive cancer detection, biomedical microscopy, nanoparticle research, Spectral and spatial resolution is a key consideration hazardous materials detection, and high- for any spectral imaging application. Hyperspectral Keywords speed waste sorting and recycling (2). instruments that are deployed for high-speed, in-line analysis require aberration-corrected sensors that can Principles of maintain imaging performance across a wide field Hyperspectral imaging Hyperspectral Imaging of view. These aberration-corrected optics are often Raman imaging peaked for optical efficiency at a specific wavelength Spectroscopy All objects reflect, absorb or emit within the spectral region of interest in order to On-line quality control electromagnetic radiation based on maximise signal-to-noise characteristics of the system. their composition. The source of this Given the demands of pharmaceutical applications, Polymorph identification electromagnetic radiation can either hyperspectral sensors cannot utilise transmissive optics

76 Innovations in Pharmaceutical Technology Issue 39

IPT 39 2011.indd 76 25/11/2011 11:22 iptonline.com

or prisms that contribute to high levels of stray light and Figure 1: Example thermal instability. of an In-Line Hyperspectral NIR Imager When deployed in-line for process manufacturing manufactured by environments, system reliability is a key parameter. Headwall Photonics; such instruments are Hyperspectral instruments (such as Headwall Photonics’ available with optical Hyperspec® In-Line Inspector system, see Figure 1) have elements peaked for the specific evolved to include not only a hyperspectral sensor core, pharmaceutical but also diagnostic routines and calibration modules to application ensure accurate, repeatable measurement and thermal wavelengths management modules that are essential for harsh imaging environments.

Applications in Pharma Production

One of the major areas of growth in spectral imaging has been industrial process monitoring and, in particular, on-line quality control applications found in pharmaceutical production. Applications relating to Images: Headwall Photonics, Inc utilising high-efficiency diffractive optics, hyperspectral and Raman sensors can be configured to offer peak optical efficiency in key wavelengths of interest across broad spectral regions, specific to pharmaceutical compounds; the major advantage here is utilisation for Additionally, when deployed in-line, hyperspectral wide area spectral analysis over a conveyor processing imaging instruments provide two significant line versus an off-line, tablet-by-tablet approach. manufacturing benefits. The first is an ability to rapidly These sensors thus present a new set of application and cost-effectively eject poor quality product or capabilities, as pharma companies are able to ramp up foreign material from the process manufacturing production with precise control over key steps in the line based on spectral discriminators. The second is manufacturing process. creation of a wealth of process flow analytics that allow a better understanding of the spectral quality Other production areas where spectral imaging holds and chemical composition of product flow at various considerable potential include the monitoring of points throughout the development and manufacturing active pharmaceutical ingredients (APIs), high-volume process. This spectral information can be used to track sampling of raw materials, troubleshooting blending product quality and safety over time from production run inhomogeniety problems, carrying out polymorphism to production run, or from one material supply to another studies, quality control of spray dry dispersion – thereby helping increase production yield and process techniques and post packaging analysis for anti- control as manufacturing is scaled to higher volumes. counterfeit verification and authentication. Hyperspectral imaging instruments greatly enhance Traditionally, capturing precise spectral information knowledge and understanding of pharmaceutical from pharmaceutical manufacturing control points processes by capturing all of the spectral and spatial has involved machine vision or single point, near attributes of material samples within the sensor field of infrared (NIR, 900 to 1700nm) spectral instruments view. When combined with spectral libraries established deployed in an off-line manner. The main limitation during the drug discovery phase and the use of of these systems is that they are only capable of multivariate analytical models, hyperspectral sensors can sampling a very small area of the overall product flow make ‘accept or reject’ decisions when deployed for ‘in-line’ and do not lend themselves to high speed production or ‘at-line’ applications. As a result, hyperspectral imaging processes. One benefit of using hyperspectral imaging allows pharmaceutical manufacturers to establish critical techniques for this type of NIR application is that it control points (CCPs) from the post-discovery phase enables high sampling rates in a short period of time, through pre-production to high-volume manufacturing. resulting in very rapid sample throughput. These hyperspectral imaging instruments can be deployed Raman Spectroscopy either as a multi-point solution or a lens-based analyser to scan large volumes of product across Raman spectroscopy measures the vibrational the production line. frequencies of various components of a molecule; these

Innovations in Pharmaceutical Technology Issue 39 77

IPT 39 2011.indd 77 25/11/2011 11:22 iptonline.com

Figure 2: Principles of light from a laser, and collecting and analysing the Raman spectroscopy, 1 2 resulting ‘scattered’ light in the following way: showing spectra of a crystalline contaminant Laser Sample in a pharmaceutical ●● Molecules in the sample ‘scatter’ the incident laser compound light, providing detailed information about the Spectrometer 3 molecular properties of the sample The Smart Bulk Excipient ●● The scattered light is collected and analysed using a Raman spectrometer to produce a Raman spectrum ●● Raman spectra are generally well resolved and rich in features enabling unambiguous identification of Trace contaminant molecular compounds detected Spectrum ●● Like traditional chemical imaging techniques that use IR wavelengths, Raman spectra can also be used 4 to rapidly identify unknown compounds ●● When coupled with high-powered microscopy, Raman imaging can be a cost-effective drug 5 development tool for evaluating the properties of drug compounds when carrying out in vitro cell- Molecule based assays (3) identified

The principles of Raman spectroscopy are shown in Polymorph/ Figure 2. A novel Raman approach – known as spatially isomer identified resolved Raman imaging – provides pharmaceutical manufacturers with numerous advantages over frequencies depend on both the bond strength and traditional approaches. Specifically, applications such mass of the bound atoms, as well as other factors such as high throughput screening of polymorphs can take as intermolecular interactions. The pattern of vibrational advantage of benefits such as higher resolution across and rotational frequencies from a molecule is highly broader sample areas, greater chemical specificity, characteristic of a given molecular species or the reduced time and improved sampling statistics. crystalline arrangement of those molecules. Whether utilised as a handheld instrument or an in- While infrared (IR) spectroscopy is essentially based line process module (see Figure 3), spatially resolved on illuminating the sample with a broad range Raman imaging has proven advantages over traditional of wavelengths of IR light and measuring which techniques, providing a greater understanding of tablet are absorbed, a Raman spectrum is obtained by composition and chemical structure. Whether used for illuminating the sample with a single wavelength of raw material screening or tablet verification, Raman imaging serves as an essential tool to improve product Figure 3: Handheld and portable Raman imagers, as well as in-line Raman instruments, performance, safety and quality. serve critical analytical functions throughout the manufacturing process Polymorph Identification Using Raman Imaging

A pharmaceutical compound can often have different polymorphic or crystalline forms, which can affect the ease of manufacture and stability of the final formulation. Therefore, detection and identification of The new platform excipient for tablets, sachets and more the different polymorphs is important throughout the • fi rst rate fi ller-binder properties due to excellent compressibility drug development and manufacturing process, so that • fast and slow disintegrating tablets (chewables, effervescents, the optimum form for eventual manufacture can be suckables, FDDTs…) selected. Additionally, during scale up and manufacture • ideal in sachets for direct oral application it is important to monitor the material being produced or dry suspensions to ensure the drug is of the correct polymorphic form • outstanding fl ow and mixing properties for inclusion in the final product. • high dilution potential and high content uniformity Raman spectroscopy coupled with microscopy is an • very low hygroscopicity and excellent stability ideal tool for characterising different polymorphic forms • GMO-free and non-animal origin

78 Innovations in Pharmaceutical Technology Issue 39 BENEO-Palatinit GmbH · Phone: +49 621 421-150 · [email protected] · www.galenIQ.com

IPT 39 2011.indd 78 25/11/2011 11:23 iptonline.com

Overview Raman spectra because the molecules of manufacturing applications and has shown itself of different polymorphic forms of ranitidine HCl experience different to be highly sensitive to the different crystalline intermolecular forces, forms of pharmaceutical compounds. Little sample resulting in differences preparation is required, making it particularly suitable Form 1 in both the frequency for polymorph identification or high-throughput raw

Intensity and intensity of the material testing. Raman spectra may be recorded from vibrational forces very small sample volumes, enabling micron-sized Form 2 observed in the Raman particles to be identified or micron spatial resolution, spectrum (4). The chemically-specific maps to be generated. Additionally sharp, well-resolved Raman spectroscopy coupled with high-powered 400 600 800 1,000 1,200 1,400 1,600 Raman shift (cm-1) features typical of microscopy has unique capabilities that make it

Example of distinguishing 1,210cm-1 pharmaceutical ideal for carrying out in vitro studies to assess the features in Raman spectra Form 1 -1 spectra also facilitate properties of drug compounds in the presence of of different polymorphic 1,186cm ® forms of ranitidine HCl Form 2 the identification of biological samples during the clinical trials stage TrackSenseTrackSense PPro Sky more subtle spectral of development. differences, which may, for example, References indicate low levels of 1. Bannon D and Thomas RJ, Harsh one polymorph in the environments Dictate Design of MeasuringMeasuring realtimerealtime shelfshelf surfacesurface temperaturetemperature andand productproduct temperature 900 1,000 1,100 1,200 1,300 1,400 Raman shift (cm-1) presence of a larger Imaging Spectrometers, Laser Focus • P Process monitoringit i andd validationlid ti one. This is exemplified World, August 2005, www. in the spectra of two laserfocusworld.com/welcomeRB. • Wireless online temperature data Figure 4: Raman polymorphs of ranitidine shown in Figure 4; this shows html?cookieName=LFWAugust2011 • No cables giving no leaks spectra from two that there are a large number of clearly resolved 2. Bannon D and Thomas RJ, Meeting Optical • Suitable for automatic loading systems different polymorphs -1 of ranitidine obtained peaks in the spectral range of 200-1600 cm , while a Demands of Next-Generation Hyperspectral • Temperature range -80°C to 400°C -1 from different zoomed-in view of the 900-1400 cm region shows the Imaging Spectrometers, Photonics Tech Briefs, • Accuracy 0.05°C (-25°C to 140°C) manufacturers characteristic features for each of the polymorphs. Oct 2004, www.ptbmagazine.com/features/ feat2_1004.html • Atex certi ed Given the lack of need for sample preparation 3. Lin J, Raman Imaging Microscopy – A Potential and the short spectral acquisition times, Raman Cost-effective Tool for Drug Development, spectroscopy is also particularly suitable for high American Pharmaceutical Review, www. throughput identity verification of raw materials. The americanpharmaceuticalreview.com/ ability to record spectra through optically transparent ViewArticle.aspx?ContentID=281 containers/windows means that materials can be 4. Webster S and Baldwin KJ, Raman tested without opening their containers, or in-line as Spectroscopy for Pharma. Part 1: Principles they are fed into the production line. and Applications, Pharmaceutical Technology Europe, June 2005, http://pharmtech. Conclusion findpharma.com/pharmtech/article/ articleDetail.jsp?id=162600 While hyperspectral imaging has long been established as a proven technology for the harsh environments of military and remote sensing David Bannon is Chief Executive Officer applications, its use in pharmaceutical manufacturing of Headwall Photonics (Fitchburg, MA), a leading designer and manufacturer operations has expanded considerably in the past few Ellab covers most of the applications shown by the vials by using high accuracy equipment. years. Understandably, a key driver has been the FDA’s of spectral imaging instrumentation. With 25 years of sales, marketing and PAT initiative to gain increased control and process engineering management expertise, he understanding at various points in the production is responsible for the successful growth and development process. Commercially available hyperspectral of the company’s strategy to develop a broad base of imaging systems can now be deployed to increase analytical instrument solutions for high performance production yields – providing a very attractive return industrial and military and defence applications. As co- on investment and short payback period, and offering founder, David led the divestiture of the optical division of the advantage of much greater, in-depth knowledge of Agilent Technologies to form Headwall. He has published the product during the manufacturing process. numerous papers and articles, and holds patents in the areas of hyperspectral and Raman imaging. Vial loading Washing Depyrogenation Filling Lyophilization Stoppering Capping At the same time, spatially resolved Raman Email: [email protected] spectroscopy has evolved to withstand the rigours

80 Innovations in Pharmaceutical Technology Issue 39 www.ellab.com [email protected]

IPT 39 2011.indd 80 25/11/2011 11:23