Near-infrared (NIR) spectroscopy has unique capabilities as compared with mid-infrared analysis because NIR does not require extensive sample preparation to obtain a spectrum. A Fourier transform (FT)-NIR instrument, a novel sample accessory, and a modified search routine for the spectral collection and identification of several different types of drug formulations are outlined, including the examination of illegal drugs.

Infrared and Raman analysis are valuable techniques for sample identification because they provide spectra specific to sample compounds. However, there are limitations to these techniques for the analysis of multicomponent samples because the multiple components often can obscure spectral peaks specific to the compound of interest. For the identification of pharmaceutical preparations, mid-infrared analysis can require extensive sample preparation to obtain spectra that can be searched against a spectral library. For either method, the presence of excipients used for tablet compounding can perturb spectral search routines, leading to incorrect identification of the primary active ingredient, which can also be present in low concentrations. This incorrect or improper identification using search databases can require further interpretation or extensive data processing of the spectral results, namely, subtraction of the excipient components. Illegally manufactured drug compounds are particularly prone to misidentification due to the use of nonstandard excipient compounds (materials for compounding of the tablets or the use of "cutting agents" used to increase the weight of the illegal drug). Near-infrared (NIR) spectroscopy, however, is a nondestructive analysis method that can be used to examine a specific sample area or provide an average of a larger sample area, depending upon the required analysis. NIR also does not generally require sample preparation for spectral collection, and the technique can be relatively insensitive for excipient materials or other contaminants. While the spectra might not be readily interpretable, chemometric analysis of the data can provide quantitative analysis as well as interpretive data with excellent precision. In recent years, NIR spectroscopy has been used widely for the analysis of biological samples and quality control and analysis of food and medical products. In particular, the diffuse reflectance technique is ideal for NIR analysis due to its simple sample handling compared to other analysis techniques. This allows rapid data collection as well as identification of drug components, similar to those methods used for process analytical technology (PAT). The use of the diffuse reflectance method in addition to chemometric processing of the data can provide a simple and rapid method for analysis and identification of various pharmaceutical and illegal drug preparations.  

A diffuse reflection accessory (JASCO model VIR-NRF-N, Easton, Maryland) was used in combination with a portable Fourier transform (FT)-NIR spectrometer (JASCO model VIR-9650) for all sample analyses.  

A drug tablet or other formulation, such as a suspected MDMA tablet, was placed directly on the sample holder and a measurement was performed without further sample preparation. An InGaAs detector was used to provide enhanced sensitivity and rapid scanning capability for obtaining the data within 10 s of placing the drug compound on the accessory sampling area. Principal components analysis (PCA) was performed to simplify positive identification of the various drug tablets.  

The various NIR spectra were analyzed using a PCA program. When the grouping of the spectral data was confirmed based upon the PCA program, a library was established using the NIR spectra. For establishing the PCA data library, 40 types of tablets were repeatedly analyzed, including 25 types of over-the-counter (OTC) pharmaceuticals, such as gastrointestinal drugs, one type of amphetamine (AP), eight types of methylenedioxymethamphetamine (MDMA; street name: ecstasy), three types of methamphetamine (MA), and three types of 3,4-methylenedioxyamphetamine (MDA; street name: the love drug). The utility of a simple identification system was examined by the investigation of the PCA searching algorithm, the calculation parameters, and a threshold established by comparison of the search results from randomly selected tablets. Various drug tablets were placed directly on the diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) sample holder (Figure 1). In the case of an extremely small tablet that was unable to be placed on the holder, the sample was placed in a test tube for analysis and placed into a powders sampling accessory (Figure 2). 

Figure 3 illustrates the PCA results for the four classes of amphetamine street drugs, including MA, AP, MDA, and MDMA. Figure 4 includes examples of the FT-NIR spectra for the four classes of illegal drugs. 

Even though the illegal drugs have similar molecular structures, sample identification is easily accomplished by the PCA algorithm as well as the spectral differences. Figure 4 also demonstrates differences in the NIR spectra where the spectral absorptions offer specific peaks depending on the tablet components. The PCA method, however, provides distinct discrimination between the various drug components due to subtle variations in the NIR spectra, even though the different drug formulations contain various excipient materials. Because NIR spectra do not provide specific discriminatory peaks like mid-IR spectra, the PCA method offers an enhanced discrimination of the various drug components, strengthening the identification of the "unknown" drug tablet. The NIR diffuse reflection system as outlined is ideal as a rapid analysis method because the tablet is simply placed on the accessory platform and the required sample measurement time is only 10 s. This method will be enhanced for sample identification as the library data is expanded with additional standard sample data. 

Additional spectra were collected from the various MDMA tablets, and the tablets were then analyzed using gas chromatography (GC) to establish the MDMA component concentration. Figure 5 demonstrates a calibration model illustrating the correlation between tablets containing MDMA and the quantitative results of these tablets using GC analysis. With a correlation coefficient of 

 = 0.966, these results demonstrate a sufficient correlation for performing quantitative analysis of the various MDMA samples. These results indicate that it is possible to analyze the amount of MDMA in illicit tablets by linking the search results of the identification program with the calibration model developed using the GC analysis. Figure 6 is an example of the program developed for the identification and quantitation of the MDMA drug formulation using the PCA search method. The results in the figure also list the amount of the active ingredient in the tablet formulation as calculated from the NIR–GC calibration model. Figure 7 demonstrates the PCA search method as applied to an OTC cold remedy, demonstrating that the NIR DRIFTS method and the PCA search program can also be used to discriminate among various prescription drugs, including the determination of counterfeit formulations. 

We have demonstrated the use of an NIR analysis method utilizing PCA discrimination for the identification of various drug formulations. The NIR system demonstrates a rapid, nondestructive analysis method for various drug tablets that can be extended to provide quantitative analysis of the drug concentration. 

 are with Jasco Corporation, Tokyo, Japan. 

Articles

The authors discuss the use of FT-NIR to identify several different types of drug formulations.

Rapid Identification of Illicit and Prescription Drugs Using FT-NIR Spectroscopy

Many analytical challenges to modern laboratories involve mixtures, whether formulated or contaminated. The infrared spectra of mixtures exhibit peaks from each component, making separation of peaks due to specific components an essential part of the analysis. This can be accomplished in two ways — spatial separation of components through microscopy or spectral separation through multicomponent searches. In either case, the final result should not just be a chemical image or spectrum, but actionable information. Further, users of varying skill levels need to obtain the same results, so consistency must be a core metric of the problem-solving tools.

Analytical and forensic laboratories are inundated with samples from many origins — contaminated product, competitive materials, or crime scene evidence. Their task often focuses on identification. In infrared, this is based primarily on comparison of the sample spectrum to library spectra. This was once done by overlaying spectra on a light box. Computers accelerated the process but did not change the fundamental approach of spectrum-by-spectrum comparisons. Both methods are still useful but experience one major shortcoming — a limited ability to handle mixtures. Digital spectral subtraction filled that gap partially, but suffered when totally absorbing peaks occurred, or when environmental changes altered peaks slightly. Because mixtures represent such a large fraction of the sample load listed above, new approaches have been sought. The two leading approaches involve spatial separation and multicomponent searching. 

Infrared (IR) microscopy permits the spatial analysis of mixtures through the collection of spectral images. For example, a tablet consists of various components pressed into a solid piece. The individual components are not distributed continuously on the micro scale, but reside in domains within the tablet. Thus, IR images of the tablet show spectral variations over the surface. In a real sense, how the image is collected — point by point or using an array detector — is not important (except as regards time). What matters is the analysis of the final image to extract information that can be acted upon. The key answers are identity, size, and distribution of the material in the tablet and relative concentrations of the components. Armed with this information, the scientist can make intelligent decisions. 

In contrast, the IR spectra of homogeneous mixtures exhibit simultaneous absorptions from the constituents. Simple searching followed by spectral subtraction works in some cases but can result in erroneous starting points and difficult-to-handle residuals. Subsequent searches might return the same result again, or might be affected by derivative-shaped bands or totally absorbing peaks. True multicomponent searching would not rely upon subtractions or other spectral processing. The result would be a list of candidate constituents and a visual display, which would again represent actionable information. 

Consistency must be an essential component of mixture analysis. Subtraction requires use of a variable factor, which can lead to poor agreement between different users. Completely removing this step is thus required to reach the consistency needed to enable less technically advanced users to obtain useable results. Further, the increased throughput demanded of analysis laboratories, coupled with the multitechnique skill sets required of the staff, mean that these tools must be automated and reliable. The following sections examine heterogeneous and homogeneous mixtures with automated analysis protocols in both micro and macro domains. 

The Thermo Scientific Nicolet iN10 MX FT-IR imaging microscope with OMNIC Picta microscopy software (Thermo Fisher Scientific, Madison, Wisconsin) was used to collect images from pharmaceutical tablets. The ultra-fast scanning of this microscope allows collection of several square millimeters in a few minutes, with a 25-μm spatial resolution. The complete analysis presented used the OMNIC Picta Random Mixtures Wizard. 

The Thermo Scientific Nicolet iS10 FT-IR Spectrometer, Smart iTR diamond attenuated total reflection (ATR) accessory, and OMNIC spectroscopy software were used to collect the IR spectrum of a pharmaceutical powder and a computer monitor cover. The thermogravimetric analysis (TGA) data were collected using the same spectrometer equipped with an internal TGA accessory. Data were exported to OMNIC Specta spectroscopy software for the multicomponent search. 

Many solid materials display granularity in the 5–200 μm range — pharmaceutical tablets, fisheye distortions in polymer sheets, and sectioned museum artifacts are common examples. Consider the fisheye example: the questions may be what contaminant caused the fisheye? How prevalent is the contamination? How is it distributed in the sample? An image alone does not convey all this information. Consider Figure 1, which shows an image of a two-component pharmaceutical tablet. The color coding represents the intensity of the signal at the scroll bar location in the bottom spectrum (about 1680 cm

) — red shows an intense signal, and blue shows a weak signal. The image communicates some distribution and identification information, but there is still untapped potential in the data. Further, if there were three components in this region, the image would not convey that information. 

Figure 2 captures an intermediate step in the wizard-driven analysis of another tablet sample. The spectral image has been digitized into a red–gray image, with red points correlating strongly to a particular library spectrum (acetaminophen, in this case). This form of image processing, based upon X-ray imaging techniques, allows additional physical information to be extracted. The analysis shows the spatial distribution but now calculates area percentages for the selected component. Cycling through the matches allows visualization of the distribution for each component. 

The whole solution — what, how much, how distributed, and where — is answered in the report shown in Figure 3. The report identifies each component, colorizes the image, and provides a calculation for the total area occupied by each component. The area percentages provide a calibrationless semiquantitative analysis of the tablet. While semiquantitative information can be somewhat useful in production QC/QA applications, in forensics laboratories this goes beyond being valuable to being essential, since only one sample might be available and no calibration standards might exist. 

Two conclusions are apparent. First, the analysis has mined the data extensively. The scientist now has the answers upon which to build action decisions. Formulations, blending, and tablet pressing modifications can be proposed based upon the identity, concentration, distribution, and domain sizes of each component. Second, the entire process occurred automatically, all the way to the final report shown in Figure 3, so any user in the laboratory would obtain the same results. This level of consistency and reproducibility greatly enhances the power and utility of IR microscopy in the laboratory. 

Homogeneous mixture analyses typically are done using transmission, ATR, or diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) methods. The ability to extract component information relies upon the additivity of the individual spectra. Essentially, the spectrum must — to a good approximation — appear as a sum of the pure component spectra multiplied by relative concentrations. In gases, this is generally true because intermolecular interactions are weak. In solids, the assumption is less true, but many materials still come close to the ideal. In liquids, intermolecular interactions, like hydrogen bonding, can cause large shifts in spectral features. Some liquids mix almost ideally, while others interact strongly — acetone and water, for instance. In the latter case, special considerations are required. In this article, only mixtures approximating the ideal are considered. 

Traditional mixture analysis begins with a first search and choice of best match. Spectral subtraction follows, and the residual is searched. There are four problems with this approach. First, the top match in the list might not be the actual best component. Figure 4 shows a pharmaceutical mixture of known composition searched in this way. The top matches are not present in the mixture, and use of these in a subtraction would lead to an erroneous beginning. 

Second, the molecular environment affects the IR spectrum strongly. Gas-phase spectra show small shifts in bandwidth and relative peak intensity with temperature, and can shift with pressure due to collisions. Subtraction in these cases leads to derivative-shaped peaks. Third, totally absorbing peaks, such as seen in many transmission spectra of polymeric films, are very difficult to remove. Most algorithms choose to ignore these peaks. Fourth, the ability to do subsequent subtractions deteriorates as the noise in the residual spectra increases at each stage. 

A much better approach is to use cumulative searching. Figure 5 shows the result of multicomponent searching applied to the same spectrum shown in Figure 4. In a sense, the same sort of microgranularity as considered in the previous section is present in this mixture, so the spectra are highly additive. The resultant three-component composite spectrum covers most, if not all, of the peaks in the IR spectrum and represents the known composition very well. Critically, the only input to the algorithm was the choice of library. 

Figure 6 shows the same algorithm applied to a heat-treated plastic component of an electronics device. The first search showed the ABS plastic, but the series of small peaks below 1400 cm

  were seen to be missing. Adding the second component yielded the flame retardant, which would allow the staff to reject this material from the WEEE/RoHS perspective (1,2). 

Two component searches are reasonably straightforward. As the number of components increases, however, the potential for interferences also increases. However, gas-phase spectra are highly additive, and four component searches have yielded excellent results. Figure 7 shows an analysis of a mixture arising from outgassing of an epoxy in a thermal decomposition (TGA) experiment. Clearly, all of the main features are assigned. 

In each of these cases, the visual agreement is very good. The match metric does not always reflect the quality of the result, due to some of the same arguments regarding spectral subtraction. Peak broadening, small shifts, or high noise will reduce this metric. However, noise has considerably less impact because there are no cumulative effects of subtraction. Totally absorbing peaks will affect the metric, but the visual comparison still shows good agreement. 

The algorithm would be even more effective if quantitative information could be gleaned from the composite percentages. However, most libraries are normalized when constructed, so the relative Beer-Lambert absorptivities between two components are lost. Hence, the fact that 20% of one spectrum and 80% of the other spectrum are required to build the composite does not translate to a 20–80 composition of the actual materials in the sample. However, the software does permit users to access spectra from the hard disk — not just data encoded in libraries — where normalization has not occurred. Use of various spectra collected in the same manner as the sample could then convert the composite percent values into good approximations for the actual concentrations, providing a reasonable quantitative analysis of the sample. As above, this analysis can be of extreme value in forensics laboratories, where it could represent the only hope for obtaining quantitative values. 

The two primary key metrics for use of multicomponent searching are met. First, the information is presented in an actionable form. The search result provides a qualitative measure of the composition of the mixture. Second, consistency is assured, because any user in the laboratory would obtain the same results. 

Analytical and forensics laboratories spend a great deal of time and effort on the analysis of mixtures. There has been a latent, urgent need for high-quality, automated tools to remove as much user-to-user variation as possible. The next-generation tools discussed here represent a major step in this analysis and provide the required consistency and automation. 

 are with Thermo Fisher Scientific, Madison, Wisconsin. 

 (1)  Directive 2002/95/EC of the European Parliament and of the Council of 27 January 2003 (Restriction on Hazardous Substances). 

 (2)  Directive 2002/96/EC of the European Parliament and of the Council of 27 January 2003 (Waste Electrical and Electronic Equipment). 

In this article, the authors discuss ways to ensure that users of varying skill levels will achieve the same results when analyzing mixtures.

Analysis of Mixtures by FT-IR: Spatial and Spectral Separation of Complex Samples

The maturity of analytical technology tends to follow two major paths: More capability with advanced features and greater performance, or smaller, less expensive, and easier to use. The former path tends to represent the evolution of today's laboratory instruments, whereas the latter path leads to instrumentation being used out of the laboratory, by a different type of end-user.

Fourier transform–infrared (FT-IR) spectroscopy is a well-established analytical technology. In the past decade, FT-IR spectroscopy has made the leap from the laboratory to at-site applications in less traditional environments. This article focuses on the next evolutionary leap that FT-IR spectroscopy is undergoing — the emergence of handheld, portable FT-IR spectrometers and the implications with respect to applications and use. 

Evolution of FT-IR Spectroscopy in Nontraditional Applications

During the 1980s, FT-IR spectroscopy became more widely accepted by personnel in QC/QA and analytical laboratories in a wide range of industries. A number of factors helped to cause this expansion of interest. Instruments became smaller, less costly, and less complicated to use. In addition to improvements in optics and electronics, the emergence of the personal computer influenced FT-IR spectroscopy. Manufacturers were able to capitalize on the economies of scale provided by using widely available computers, and a broader range of users were able to adapt to these systems rapidly because of the intuitive nature of the software.  

Another important driver for the widespread adoption of FT-IR spectroscopy was the development of sampling technologies that opened up a much broader range of applications. IR spectroscopy sampling moved from the classic preparation in potassium bromide pellets and Nujol mulls, to methods that allowed for the examination of neat samples with little or no preparation. These latter methods included internal reflectance (attenuated total reflectance [ATR]), diffuse reflectance, and FT-IR microscopy. 

Thus, FT-IR spectroscopy saw rapid growth as a valuable, routine laboratory method used by personnel who required quick information and answers. 

During the 1990s, FT-IR spectroscopy began to be used in applications and environments that were considered more nontraditional. For example, process FT-IR spectrometers, which could withstand the rigorous environment imposed by the manufacturing plant, were introduced. Concurrently, FT-IR spectroscopy began to be "rediscovered" by organic chemists and chemical engineers to aid in the development of complex chemistry in process development and process engineering facilities. For example, chemical engineers found that the integration of in-situ FT-IR spectroscopy and reaction calorimetry provided critical information about reaction heat flow, composition, and mass balance.  

To support these latter applications, new FT-IR sampling technology was required. In the mid-1990s, the diamond ATR and diamond ATR probe were developed. This technology permitted FT-IR internal reflectance spectroscopy to be carried out in chemically aggressive solutions that would have destroyed conventional ATR crystals. Due to its inertness and resistance to scratching, the diamond ATR sampling interface became a widely used sampling technology that made FT-IR amenable to an extremely broad range of applications. 

The transition of FT-IR spectroscopy from the traditional laboratory to manufacturing and development laboratories was well under way during this period. It also became clear that rugged, smaller, easier-to-use, and less costly systems had great potential and appeal. 

Starting in 2000, portable FT-IR systems were developed for applications such as forensic investigation of illegal drug laboratories; for assisting United Nations weapons inspectors in Iraq in their search for weapons of mass destruction; and for helping emergency responders to identify a wide range of potentially hazardous substances. These applications proved that portable FT-IR spectroscopy could be deployed effectively in challenging environments by nontraditional users. 

The success of these portable FT-IR spectrometers for dedicated, nontraditional applications led to the next step in this branch of FT-IR evolution, the development of systems that are even smaller, lighter, easier to use, and less costly — the handheld FT-IR instrument. As an example, the Exoscan handheld FT-IR system (A2 Technologies, Danbury, Connecticut) was developed for surface analysis of metals and composite materials in cooperation with companies involved in commercial aviation manufacturing. 

Considerations in the Development of a Handheld FT-IR Spectrometer

The operating goal for a handheld FT-IR spectrometer is to employ the system at the analysis site. This is particularly true when the object to be analyzed is too large to bring to the laboratory or too valuable or too difficult to remove a sample. This implies that a handheld FT-IR spectrometer will likely be used in environments where ambient conditions are a significant factor. Thus, the instrument design must take variables in temperature, humidity, and vibration into consideration. Equally important is the nature of the sample to be analyzed, the type and responsibility of the operator, and how the system actually will be used. The latter point is critical because, for example, a handheld FT-IR spectrometer must demonstrate stability and reproducible performance in virtually any physical orientation.  

Obviously, the size of the optics is a major consideration, and there are tradeoffs. If the optics are too large, the overall package can be unwieldy. However, performance might be compromised if the optical system is too miniaturized. The type of application for which the system will be deployed is an important driver for the decision about the tradeoff between performance and size. For surface analysis of composites and metals, the system must perform at a level equivalent to or better than traditional analytical FT-IR spectrometers. In the Exoscan system, this is accomplished by minimizing the internal optical path while maintaining a large beam aperture, thereby ensuring that the optical throughput of the system meets the requirements of the application. The instrument's optical system and electronics weigh approximately 2 lb, and the overall system weight, including internal battery and data system, is approximately 6 lb.  

Thus, the handheld spectrometer must be rugged, insensitive to vibration, capable of handling temperature fluctuations, and not affected by motion or position. System optics must be aligned permanently, requiring no external adjustments, and designed so that the overall system remains stable, with consistent performance.  

A handheld FT-IR instrument will likely address a wide range of sample types. The samples might be liquid, solid, or semisolid. They might be highly colored, opaque, reflective, or highly absorbing. But in all cases, the nature of the analysis implies that no sample preparation is required. The sampling interface must be versatile, yet the sampling optics must be integrated to provide maximum throughput with minimum impact on the overall system size. 

Because of the applications that the Exoscan instrument addresses, two different sampling systems were developed. For hard surfaces or surfaces that are reflective, such as metal and composites, an external reflectance interface is used. For samples that are slightly softer or nonreflective, such as paints or polymers, a diamond internal reflectance interface is used. These two interfaces are user interchangeable with bayonet-type mounting to eliminate the requirement of realignment when the sampling interfaces are exchanged. 

A handheld FT-IR spectrometer will be used by at least three types of users, all with different skill sets and different responsibilities. One group of users will be scientists who are responsible for developing and validating methods that are to be deployed on the system. These individuals are well trained in spectroscopy and require software that enables them to develop a variety of methods including sample identification methods, quantitative methods using univariate or multivariate algorithms, and methods that define whether an article is acceptable or unacceptable for use. The methods might be executed by personnel who likely will be untrained in spectroscopy and are using the handheld FT-IR system to acquire specific information. The type of information that these users require might be presented in a "pass–fail," "yes–no" format or as a specific value that is indicative of the performance of a sample. These users require the system to operate intuitively, with simple push-button operation. The information that the operator obtains must be easily transferable to another type of user, the supervisor, who is responsible for certifying or otherwise validating information from the system as well as verifying that the FT-IR system is operating correctly. 

The potential range of applications for a handheld FT-IR spectrometer is quite broad. Some of these applications include: analysis of materials such as polymers and composites; determination and quantification of surface contaminants and surface quality; identification and measurement of coatings, including paints, primers, and films; and general identification of a broad range of liquids, solids, and semisolids that are used in the manufacture of chemicals and chemical-based products. The following examples are applications that a high-performance, handheld FT-IR spectrometer can accommodate. 

Analysis of Damage to Composite Materials Caused by Heat, Environmental Oxidation, and Chemicals

The Exoscan instrument was developed in collaboration with companies in the aircraft industry, and the requirement of analyzing black, medium modulus, graphite fiber resin composites was one of the drivers for the performance and sampling interface design.  

Because composites are organic materials, they are susceptible to environmental stressors that potentially can weaken the composite material. For this reason, there is significant interest in nondestructive analyzers that can determine the condition of in-service composite materials. 

Analyzing composites via FT-IR requires some special considerations. In fiber-rich, resin-poor regions of the composite, the amount of IR radiation being reflected by the sample is highly attenuated by the carbon fiber, so it is very important that the performance of the spectrometer is optimized. An external reflectance sampling interface is required, enabling the spectrometer to analyze the resin component of the composite for molecular changes caused by exposure to a variety of stressors. 

As an example, the system was used to analyze cured composite samples that were heated for 1 h at 25 °F intervals starting at 300 °F (Figure 1). The IR spectra of the heat-treated composite samples show increasing growth of an ester–perester type carbonyl species, and additional ketone–carboxyl groups as the thermal treatment temperature increases. These changes are indicative of significant oxidation, which results in weakening of the composite matrix. The IR absorbance bands associated with aromatic polyethers (PEEK) and polysulfones (PSF) are observed to diminish greatly in the 350–500 °F temperature range. These PEEK and PSF composite resin bands are observed to inversely correlate quite well with the treatment temperature. The loss of these functionalities is indicative of a weakening of the resin bonds, which in turn affects the overall strength of the composite. 

Contaminants on Surfaces and the Effect on Bonding and Coating

Successful bonding and coating of metal or composite materials is highly dependent upon how the surfaces to be joined or coated are prepared and cleaned. With FT-IR spectroscopy, residual contaminants can be detected and identified, and the levels present can be measured.  

Contaminants can be entrained in primers that are applied to surfaces. Oil from fingerprints and machinery (Figure 2), amount of water present in the primer, and humidity of the ambient environment can affect the primer performance. These contaminants are readily detected by FT-IR spectroscopy. Also, silicone mold release is important in composite fabrication, but residual silicone is a troublesome contaminant with respect to bonding and coating processes. Fortunately, silicone contaminants, from a variety of sources used during the production of metal and composite parts, are detected by FT-IR spectroscopy. 

Metals such as aluminum often are given an oxide coating via anodization to minimize corrosion. There are different methods for anodizing metal, and each produces different chemical structures on the metal surface and different physical properties. Handheld FT-IR spectroscopy readily elucidates these different anodized surfaces to ensure that the proper surface treatment has been applied, and provides the amount, thickness, and uniformity of the anodized coating.  

As an example, five samples of 2024 aluminum were treated with borate sulfuric acid anodization and measured with the handheld FT-IR instrument (Figure 3a). The aluminum oxide surface concentration ranged from 66 to 374 mg/ft

. The strong aluminum–oxygen stretching band observed at 1128 cm

  shows good correlation with concentration of surface oxide layer (Figure 3b). 

As a result of ambient temperature and humidity conditions, unintended oxidation of metal surfaces can occur as well. This thin layer of oxidation can significantly affect follow-on steps required to coat, paint, or prime a surface. Handheld FT-IR spectroscopy can detect this unwanted oxidation and provide a measurement of the amount present. 

Paint, Primer, and Adhesive Identification, Homogeneity, and Thickness 

The design documentation for many aircraft specifies paints of a particular composition. These paints are selected for durability, weight, weather resistance, and ultraviolet resistance. In addition, primers are specified for their particular bonding characteristics. Often, articles will arrive either primed or painted from outside vendors. Nondestructive FT-IR analysis can be used to verify the composition of these paints and primers on full parts and eliminate the need to disassemble or otherwise destroy expensive parts for first article inspection. 

For example, epoxy paints and primers, as well as urethane paints, were identified on aluminum parts with the handheld instrument. Though libraries can be created using the instrument's software, in this particular study a library generated using a conventional laboratory FT-IR system was searched. The samples of aircraft epoxy paint were measured using the diamond ATR sample technology, and the spectra were collected in approximately 4 s. Figure 4 shows an excellent match between the unknown paint measured by the handheld instrument and the library spectrum measured on the laboratory FT-IR spectrometer, verifying that the correct paint was used. 

By analyzing the molecular structure of paints and primers, FT-IR can differentiate different paint and primer formulations. A handheld instrument with an onboard library can verify that the proper material has been applied to a surface as well as indicate the uniformity and thickness of the surface coating. 

This article has provided a perspective on the evolution of FT-IR technology over the past 30 years. The scientists and engineers who have been involved over this period have participated in the growth of a highly valuable and interesting technique. Each person likely would write the history differently, based upon their own experience, but we would all likely agree on the outcome, that FT-IR spectroscopy continues to be an exciting technology with great growth potential driven by new, commercially valuable applications. The advent of handheld FT-IR spectroscopy now offers even greater opportunity for expansion into new applications, markets, and industries.  

 is Vice President, Business Development, with A2 Technologies, Danbury, Connecticut.

This article focuses on the emergence of handheld, portable FT-IR spectrometers and the implications with respect to applications and use.

Handheld FT-IR Spectrometers: Bringing the Spectrometer to the Sample

Vibrational spectroscopy provides abundant analytical information on materials, which can be utilized in product development, material evaluation, and failure analysis. For technical reasons, measurements of direct vibrational absorption (FT-IR) developed more rapidly as a reliable tool in this area. Again for technical reasons, Raman scattering measurements have now achieved analytical quality and convenience. However, while the two techniques both measure vibrational spectra of materials, the spectra are quite different — different functional groups tend to be stronger in one measurement rather than the other; the two techniques provide different information. Polar functional groups provide strong IR absorptions, whereas aromatics and double bonds provide strong Raman signals. Consequently, the ability to acquire Raman and IR signals on the same sample provides a synergism over the sum of the individual measurements.

All that is said in the above paragraph can be repeated for spectroscopic microscopy (or microscopic spectroscopy, depending on your preference). But here the interest is in the microscopic heterogeneity and distribution of the various components in a sample. The ability to record Fourier transform (FT)-IR and Raman spectra sequentially on the same sample, using the same nosepiece so that the sample does not have to be disturbed, means that the synergism of the two spectroscopies can be used to characterize the nature of the samples. The Raman and the FT-IR spectra can be measured on exactly the same spot; well, not exactly because the limiting spot size depends on the wavelength of the probe (about 10–30 μm for FT-IR and 0.5 μm for Raman, depending on the optics). 

Then, after anyone starts to use a microscope for spectroscopic measurements, the immediate interest is in measuring the components' distribution — that is, collect a map. Because FT-IR and dispersive Raman instruments normally are manufactured by different companies, the software interface for the two systems does not work together. However, the results that will be shown here were acquired on a LabRAM HR high-resolution Raman microscope equipped with a Smith's Detection interferometer mounted on top of the microscope, all controlled by LabSpec software (Horiba Jobin Yvon, Edison, New Jersey) (1). 

There are different methods of implementing vibrational mapping. They essentially fall into two categories, with hybrid variations in between: illuminate a point on the sample, collect a spectrum, then move to the next point; and illuminate the entire region of interest, take a snapshot at one wavelength, and then move to the next wavelength. In both cases, a hyperspectral cube is created: for each 

 point on the sample, there is a spectrum with intensity as a function of wavelength. The chemical map is then constructed using features of the spectra. Plotting intensities of selected bands is the simplest method, but others are also available. The modeling function is useful in cases where the spectra are complicated and unique bands are difficult to identify. In this case, each data point in the map is decomposed into a linear combination of "pure" spectra that are selected from the data set, acquired from pure samples, or calculated by the software algorithm. It is also possible to "bandfit" the spectra to Gaussian–Lorentzian functions and then plot characteristics reported by this program, which include peak position, peak width, peak amplitude, and integrated area. 

The following example was taken from a brief study that was done on bone tissue. While the FT-IR microscope can accommodate both an attenuated total reflectance (ATR) accessory and an all-reflecting objective (ARO), it is the ARO that is used for mapping studies. Figure 1 shows the microscope nosepiece with ARO objective mounted, and a schematic of the second generation of the objective. Figure 2 shows the spectra of the bone and the polymer matrix. 

A captured image of the mapped region is shown in Figure 3. The sample is unstained so the contrast is entirely determined by the materials themselves. Major features are discernable, but detailed information on the molecular content in the image field can only be inferred. On the other hand, Figure 3 shows the FT-IR image using the intensity of the phosphate band of the bone just above 1000 cm

, and the polymer's carbonyl band near 1735 cm

In reality there is more information in these measurements than what is shown in Figure 5. The spectrum of the bone is composed essentially of the phosphate mineral (major band at 1036 cm

) and protein (amide I and II bands near 1650 and 1550 cm

), and the phosphate is carbonated (indicator band at 888 cm

). Images of each of these species can also be constructed, as shown in Figure 4. Not surprisingly, all three chemical components in the bone overlap, but careful inspection of the images indicates that the relative contributions of mineral to protein vary over the image.  

Raman Microscopy of the Same Bone Specimen

As in the FT-IR mode, it is simple to distinguish the bone from the matrix by Raman spectroscopy. Figure 6 shows the spectra of these two components and Figure 7 shows the positive–negative images of the two species constructed by using the intensity of the bone phosphate stretch at 960 cm

As with the FT-IR, there are spectral characteristics of the bone that show the carbonate (approximately 1080 cm

), and phenylalanine in the protein (approximately 1000 cm

). Figure 8 shows the ratio of the phosphate (960 cm

) intensities. In this image, the mineral content of the bone is highest in the center, suggesting that as the bone grows out from the center, first protein is deposited and then the mineral.  

Note that the two sets of results (FT-IR and Raman microscopes) shown here yield similar information — distribution of bone and polymer embedding medium. The question is why would one bother with the two techniques? It is actually possible to extract more information from both spectroscopies. The group of Rick Mendelsohn at Rutgers University in Newark, New Jersey, has been doing FT-IR profiles of bone for years (2). By applying sophisticated analytical methods, they have been able to extract chemical and structural information from spectral variations. 

What the maps shown above do not show is the trade-offs between the two techniques. In general, FT-IR results are acquired much more rapidly than Raman, as was the case here. But a Raman microscope provides higher spatial resolution and spectra with sharper bands. At least two groups are active in utilizing Raman microscopy to characterize bone (3,4), also extracting information such as mineral to matrix ratio, degree of carbonation, and protein structure, which can be related to disease states and to aging. 

With the maturation of FT-IR and Raman microscopies, there are now several instruments on the market that integrate the two measurements on a single platform, enabling synergistic studies of many materials. As an example, some simple maps of bone have been shown. 

 are with Horiba Jobin Yvon, Edison, New Jersey.

(1)  F. Adar, A. Whitley, and J. Reffner, LabRAM IR with Same Spot Technology, Pittcon 2002. 

(2)  E.P. Paschalis, E. DiCarlo, F. Betts, P. Sherman, R. Mendelsohn, and A.L. Boskey, 

(3)  C.T. Tarnowski, M.A. Ignelzi, W. Want, J.M. Taboas, S.A. Goldstein, and M. Morris, 

(4)  O. Akkus, A. Polyakova-Akkus, F. Adar, and M.B. Schaffler, 

The authors discuss the use of instruments that combine FT-IR and Raman microscopy on a single platform, enabling synergistic studies of many materials.

Integrated FT-IR and Raman Imaging on the Same Microscope

The rapid and accurate identification of microbial organisms is increasingly important in several critical areas of interest. For example, the identification of foodborne pathogens plays a critical role in determining the source of food poisoning, such as the recent outbreak of salmonella from tomato consumption in the United States. Many microorganisms have highly specific infrared absorption signatures and are suitable for Fourier transform–infrared (FT-IR) spectroscopic analysis. FT-IR spectroscopy can generate spectra of microorganisms without the need for extraction, amplification, labeling, or staining and can provide results in much less time than traditional identification methods. The key to success in identifying microorganisms down to the strain level with FT-IR spectroscopy is achieving excellent reproducibility in sample growth conditions, correct preparation of the biofilms, and subsequent collection of the infrared spectrum.

 are causing increasingly widespread outbreaks of illness across many states and nations. Each year, millions of people in the United States experience illness caused by a foodborne microorganism. Securing the safety of the food supply requires the ability to detect and classify contaminates before a product is placed into the food supply. Current methods of identifying microorganisms require resuscitation of pathogens using a series of nutrient media transfers beginning with nonselective enrichment media and moving to a selective medium. This technique generally takes several days to complete, so by the time a contamination has been detected, the products usually have been released to market. This long lead time can compromise public health and lead to expensive and reputation-damaging recalls that are sometimes necessary in the interest of public health (1).  

There is also a growing need for rapid identification of microorganisms in patients in order to promptly prescribe the correct treatment regimen. Traditional commercially available identification systems are based on the microorganism's nutritional and physiological characteristics. A pure microbial culture and large inoculums are required to obtain results with these methods. The typical turnaround time is 1 day for completion of the process of extraction, amplification, labeling, and staining required for delivery of identification results. This delay can have adverse consequences. One study showed that mortality rates of patents who do not initially receive the correct therapy for a microbial infection is 30% to 60% higher than the group that receives the appropriate therapy promptly (2).  

Development of FT-IR Spectroscopy for Microbial Identification

Fourier transform–infrared (FT-IR) spectroscopy has been used in industrial applications such as chemical and pharmaceutical manufacturing and quality control for decades to provide positive fingerprint identification of materials, but only more recently has been applied to the challenge of microbial identifications. The development of an FT-IR spectroscopy–based method for the identification of microorganisms was carried out by researchers at the Robert Koch Institute in Berlin in collaboration with Bruker Optik GmbH between 1986 and 1989. The prototype of a dedicated FT-IR system for microbial characterization was then tested by several microbial laboratories in a joint pilot study between 1992 and 1994 (3–5). 

FT-IR microbial identification is based on the fact that each molecule's functional groups absorb infrared radiation to generate a characteristic absorption or transmission spectrum that is rich in information and unique to that molecule. Spectra can be analyzed or searched against libraries of reference materials to positively identify unknown materials. Just like human fingerprints uniquely identify their owner, IR spectroscopy provides a spectral fingerprint that uniquely identifies a chemical compound. In the field of microbiology, an FT-IR spectrum reveals a cell's fingerprint, which reflects its biochemical composition including proteins, lipids, DNA and RNA, and carbohydrates. The method's high sensitivity and selectivity make it possible to identify microorganisms even down to the strain level.  

The cell material of a pure single colony is spread on an agar plate using a spatula to generate a confluent growth, as shown in Figure 1. The optimal incubation time is typically 16–24 h, and the most common growth temperature range applicable to bacteria is 35–37 °C. 

Sample material can be harvested directly from the middle of the cell lawn using a calibrated 1-mm diameter platinum loop. A calibrated bacterial loop should be used to avoid transferring agar. The necessary amount depends on the organism and the consistency of the colony growth, but typically corresponds to one or two full to heaping loops. The sample material should only be taken from the confluent growth regions of the colonies in the third quadrant of the agar plate. This technique captures a mixture of cells in a very reproducible mixture of different growth phases. The material is transferred and suspended in water in the reaction vessel to create a homogenous biofilm. Normally, shaking the reaction vial is sufficient to suspend the bacteria, but in some cases a vortex or even weak ultrasonic treatment might be necessary. 

An aliquot of the suspension is then transferred to the sample position of the microplate with a pipette and dried to a biofilm for the final measurement. The suspension should be placed to cover the sampling region completely so that the infrared beam cannot penetrate the sampling area without passing through the sample. The sample is placed over silica gel or Siccapent inside a closed desiccator at 40 °C. A vacuum of 4–7 kPa (30–50 Torr or 40–70 millibar) is applied for 10–20 min depending on the number of samples and the condition of the desiccant to dry the sample to a transparent bacteria film. The desiccator should be vented not more than 5 min after the last sample is dry to prevent the sample films from tearing. Testing might be necessary to determine drying conditions that can provide high spectral quality. 

The cuvette can then be loaded into the spectrometer sample compartment. Typical analysis takes about 15 s. 

The evaluation of the microorganism spectra should be undertaken using the first derivative of the resulting spectra. Broad overlapping bands characteristic of microorganism spectra become better resolved by taking the first derivative, as shown in Figure 2. Taking the first derivative of the spectra also provides a simple baseline correction. The intensity of the bands in the first derivative spectra is approximately 100 times smaller than in the original spectra. Measured spectra can be evaluated using either cluster analysis or an identity test report.  

Cluster analysis helps identify similarities between the spectra of microorganisms under analysis and the spectra of known species and can be most useful for differentiating spectroscopically similar strains. The wavenumber regions, weighting factors, and mathematical methods used for cluster analysis must be specified. Two very popular cluster analysis algorithms for microbial identification are average linkage and Ward's algorithm. A dendrogram is a tree diagram frequently used to illustrate the arrangement of the clusters produced by a clustering algorithm. In the average linkage method, the fusion values in the dendrogram correspond at least approximately to the original distance values in the spectra, making the dendrograms relatively easy to interpret, as illustrated in Figure 3. 

Established and reliable reference techniques have been used to develop spectral libraries of microorganisms. These libraries are important for the identification and classification of microorganisms. The following procedure can be used to create a library with fixed wavenumber ranges. Repeat measurements are carried out over time of independently prepared strains of the sample cultivated under identical conditions in different batches of the same medium. A valid spectral database should cover a high number of different species with a high number of reference strains for each species. The composition should be well balanced, and all reference strains should be identified carefully by independent methods. 

The strain typical spectra are partitioned randomly into two approximately equal size data sets — the original data set and the test data set. The re-identification rate (the percentage of the test data set correctly identified at the strain level) is calculated based on each wavenumber region, one at a time using the original data set as the library and the test data set as samples. The six wavenumber ranges shown in Table I have proven to be particularly useful for the identification of microorganisms. These frequency ranges have been selected based on the empirical values of selected organism groups. For other organisms, different regions might have superior predictive value.  

The wavenumber ranges are selected to minimize the number of incorrectly identified spectra. They are put together in combinations with each range having a weighting of the re-identification rate. These combinations of wavenumber domains are tested, and the optimal combinations are selected for the spectral library. The entire procedure is repeated for validation purposes, except that the original data set and the test data set are interchanged. Finally, the spectral library is tested on a data set of spectra collected from microorganisms not included in either test set to ensure that no systemic errors have been introduced.  

Strains should be organized into sensible classes such as species, genus, or other systematic grouping, so that unknown species can be clearly assigned to a known group. An automated process can then compare the spectrum of an unknown microorganism against all of the spectra in a library. This process can incorporate the specific wavelength range used, weighting factors and mathematical techniques. Identification results are summarized in table format showing the identified strain or species. Additionally, each sample has a detailed hit list. Statistical methods are used to express the probability that the identification is correct and determine the specific limit values for each group. The reporting of the identification probability is required under GMP and GLP guidelines. Each measurement of known strains increases the reference database, which in turn increases the precision of the determination of the individual microorganisms. 

In the dendrogram shown in Figure 4, the spectrum of the unknown bacterium shows high similarity to the 

 spectra. The small values for hit quality for the two 

 species in the reference database confirm the identification as a 

 is below the threshold value for this group. This indicates that the unknown strain belongs to the 

 species. The spectra should be compared to a database that contains only spectra of 

FT-IR spectroscopy is a whole-cell fingerprinting technique that provides high sensitivity, subspecies specificity, and applicability to all microorganisms. Further, it can be used to identify and classify microorganisms by genus, species, and strain utilizing the well-established mathematical methods of cluster analysis and quantitative searching of reference data sets. The FT-IR analysis can be automated and has low associated consumable costs. The mean time for identification and classification is about three times shorter than currently used methods, which can have an important effect on early detection and management of an outbreak of potentially hazardous microorganisms. 

 is Applications Manager with Bruker Optics, Inc., Billerica, Massachusetts.

(2)  E.H. Ibrahim, G. Sherman, S. Ward, V.J. Fraser, and M.H. Kollef, 

(3)  D. Naumann, D. Helm, and H. Labischinski, 

(4)  D. Helm, H. Labischinski, G. Schallen, and D. Naumann, 

(5)  D. Helm, H. Labischinski, and D. Naumann, 

The author discusses the benefits of using FT-IR to identify microbial organisms.

Microbial Identification Using FT-IR

The use of thermo micro methods for the evaluation of active pharmaceutical ingredients and excipient interactions enables direct and rapid analysis of microscopic samples. The value of this capability is significant at the earliest stages of drug development when time is important and samples available for testing are limited. Applying infrared spectroscopy to these microscopic sample preparations extends the value of results generated and provides molecular analysis of physical and chemical interactions.

The use of infrared (IR) spectroscopy for chemical identification and evaluation is limitless. Whether the challenge faced is as simple as identifying a raw material before receipt into a manufacturing plant or as complex as monitoring a multispecies reaction to determine endpoint, IR spectroscopy can be applied effectively to solve a variety of problems. The development of a miniaturized interferometer revolutionized the application of this well-established technique, effectively making it possible to link an IR spectrometer to standard light microscopes. Unfortunately, regardless of their usefulness in a variety of applications, IR microscopes are not always used to their fullest potential.  

Instances commonly arise where the need to collect an IR spectrum from a defined area within a microscopic sample is critical. For example, when analyzing cross-sections of multilayer laminates, effective analysis would not be possible without the ability to visualize each layer and collect spectra from discrete locations across the sample. Or when analyzing specific crystals within a mixture, the ability to microscopically differentiate and select a single particle for analysis is necessary. Drug chemists frequently rely on this capability when isolating illicit drugs for analysis from among complex cutting-agent matrices.  

A field of study that receives a significant amount of attention for a variety of reasons is the analysis of solid-state form of active pharmaceutical ingredients (APIs) and the compatibility of these molecules with excipients used in drug-product formulations. When applied within this field, the value of the IR microscope can be realized. Not only is it possible to evaluate distinct particles to evaluate changes such as crystalline or amorphous structure, hydrogen bonding, and water or solvate bonding, but it is also possible to apply thermo micro methods (1) to evaluate the nature of the interactions between different compounds in a formulation to assess component compatibility.  

Thermo micro methods have long been applied to the characterization and identification of unknown substances. Using a microscope for observation, a sample of unknown composition is heated under a glass microscope slide alongside a "standard" of known melting temperature. The interaction of the two substances is monitored, and the effect each compound has upon the other is evaluated. The simplest interaction resulting from this type of fusion preparation is formation of a eutectic. This is a uniform fine-grain mixture of the two components. It has a single melting temperature that is always lower than the melting temperature of either starting material. Comparison of the eutectic melting temperature with tables of known eutectics enables an identification of the unknown. Figure 1 shows a fusion preparation that formed a simple eutectic. At the time this image was collected, the sample was heated to its eutectic melting temperature. The eutectic melt is observed along the zone of mixing between the two starting materials. 

Using an IR microscope, it is now possible to collect spectra from along the zone of mixing as well as from across the fusion preparation to perform molecular identification of all compounds present. Application of IR spectroscopy to thermo micro methods is ideal for the evaluation of API–excipient interactions. Sample amounts are microscopic, analysis time is minimal, and in addition to the ability to observe physical and chemical interactions directly, spectral analysis of eutectics and other complexes formed (such as co-crystals or addition compounds) is performed easily. The value of this type of information early in the drug development process is substantial because sample quantities of new molecular entities usually are limited and the need to maximize testing efficiency is crucial. In addition, the use of these methods makes it simple to create composition diagrams that can be used to evaluate the effect manufacturing conditions such as heat will have upon the API and excipients. These diagrams can be used as guides when establishing formulations and manufacturing procedures. Figure 2 shows the composition diagram of a simple eutectic. 

A eutectic is a uniform physical mixture of two substances and behaves as if it is a single phase. Therefore, the chemical composition of a eutectic is fixed and can be measured easily by performing a series of experiments using a hotstage microscope. When performing these experiments, different concentrations of the two components are mixed intimately. These mixtures are then melted to the eutectic temperature to determine at which concentration only a melt is observed (no excess crystals of either component A or component B are present). This concentration is the concentration of the uniform eutectic. Using IR microspectroscopy, this process is simplified. In addition, the IR spectrum provides information about the sample that is not obtainable using hotstage microscopy. For instance, because the eutectic is a physical mixture of two components, the IR spectrum of the eutectic is a mixture of the IR spectrum of each component. There should be no additional absorption bands present in the eutectic spectrum, and only minimal shifts in bands resulting from chemical interactions between the two components. Conversely, if the IR spectrum shows additional or different absorption bands or significant shifts in band position, it is likely that the interaction between the two components is chemical in nature (that is, addition compounds or co-crystals have formed). When a new compound is formed, the new molecular entity might form a eutectic with pure component A on one side and pure component B on the other. A composition diagram of a two-component system that forms an addition compound with eutectics on either side is shown in Figure 3. A temperature-composition diagram is a convenient means of expressing the melting point as a function of composition of all solid phases present in two- or three-component systems (2). The value of the IR microscope to collect spectral information is important because it provides molecular information about the individual components as well as how these components react when combined. 

To evaluate the ability of the IR microscope to evaluate these types of fusion preparations effectively, a series of samples known to react with each other and form simple eutectics were prepared and analyzed. Application of the IR microscope to these samples is invaluable because the spectral results provide a molecular understanding of the nature of the compatibility of the different components.  

Five different eutectic samples were prepared using the fusion-preparation method. Each sample was analyzed using a Smiths Detection IlluminatIR II infrared microscope with a Linkam LTS350 hotstage (Danbury, Connecticut). For each sample, at least five spectra were collected along the zone of mixing (that is, along the eutectic, at a temperature below the melting temperature of the eutectic). All eutectics were analyzed in their solid form. In addition, a spectrum was collected from pure component A and pure component B of the same fusion preparation. Spectral libraries of linear combinations of pure A and pure B in 5% increments were created for each sample. Each individual spectrum collected from along the zone of mixing was searched against the library prepared for that two-component mixture. The spectral match with the best hit quality was determined. The average percent composition and standard deviation of the higher-concentration component, for the spectra collected along the eutectic, was determined. The eutectics prepared are detailed within Table I. 

Table I: Two-component mixtures used to prepare eutectics 

Spectra from the azobenzol–benzil eutectic and the phenacetin–benzanalid eutectic were collected using reflection absorption microspectroscopy with an analysis spot size no greater than 100

  at each point analyzed. The remaining three eutectics were analyzed using attenuated total reflection (ATR) microspectroscopy with an analysis spot size no greater than 42

In addition, a fusion preparation of trinitrofluorenone and mononitro-napthalene was prepared. This two-component mixture is known to form an addition compound with a melting temperature of 163 °C. An IR spectrum was collected from this addition compound and compared with the IR spectra collected from trinitrofluorenone and mononitronapthalene.  

Each IR spectrum collected along the eutectic of each sample was consistent with the spectra collected from different locations of the same eutectic, indicating that the eutectic is chemically consistent throughout. Figure 4 shows an image of a typical eutectic as it appears at temperatures below the eutectic melt. In this image, the crystal front of the pure benzanalid phase is seen at the top left corner of the field. The crystal front of the pure phenacetin phase is seen at the bottom right corner of the field. The zone of mixing where the interaction between the two phases occurred is across the image. Five locations from which IR spectra were collected from this sample are highlighted in yellow. Note that the eutectic formed is a finer-grain phase than either the crystals of benzanilid or phenacetin. 

Data collected from each eutectic is summarized and displayed in Table II. These results show that the chemical composition of each eutectic, as expected, is consistent from location to location. 

Table II: Summary of analysis of eutectic composition 

In addition, a review of the IR spectral data shows that the average spectrum collected for each eutectic is a mixture of the spectra from each of its pure components. In four of the five examples, no additional bands or shifts were observed in the eutectic. However, for the eutectic of salophen and dicyandiamid, shifts in bands were observed across the spectrum. The exact nature of these shifts is not known. It is likely that although the eutectic is assumed to be strictly a physical mixture of the two components of the mixture, for this sample there is a measurable degree of chemical interaction between these two components in the eutectic phase as evidenced by differences in the IR spectra. Figures 5 and 6 show the average eutectic spectrum of azobenzol and benzil. No additional bands or significant band shifts are present in this eutectic spectrum. This type of spectral comparison was typical of all of the eutectics formed except for the spectra from the salophen–dicyandiamide eutectic. Figures 7 and 8 show the average eutectic spectrum of salophen and dicyandiamide. While the spectrum is an overlay of the two spectra, and even though a single-phase melt is observed at the eutectic melt temperature, the IR spectra indicate that there is some interaction between the two components that is likely chemical in nature. 

The fusion preparation of mononitronapthalene and trinitrofluorenone is shown in Figure 9. Mono-nitro-napthalene is on the left side of the image and trinitrofluorenone is on the right side of the image. The addition compound is along the center of the image, is dark brown in color, and is highlighted with a yellow arrow. The spectra collected from the addition compound as well as from pure mononitronapthalene and trinitrofluorenone are shown in Figures 10 and 11. These spectra show that the molecular composition of the addition compound is different from each starting component and is not simply a physical mixture of the two. Knowing the chemical structure of the starting compounds, it is possible to evaluate what functional groups of the molecule are changing chemically. The ability to perform this type of analysis is critical when it is necessary to understand the nature of changes occurring. 

The use of thermo micro methods enables an evaluation of the melting characteristics of each compound formed during interaction. While these methods have historically relied upon melting behavior, the addition of IR microspectroscopy increases the value of the information obtained significantly and is an ideal tool for evaluating API and excipient compatibility. 

Application of the IR microscope to thermo micro methods significantly increases the valuable information obtained from the analysis of APIs and excipients. Historically, these methods were used for the identification of unknowns by visually observing component interactions and melting behaviors. Using the IR microscope, it is now possible to understand, on a molecular level, the nature of interactions between APIs and excipients. The added benefits of rapid analysis times and needing only microscopic sample amounts make these techniques ideal for use during the earliest stages of the drug development process. At this time, knowledge of API and excipient interactions can effectively be used to guide formulation development and the manufacturing process. 

 is with Smiths Detection, Danbury, Connecticut. 

 is with Purdue Pharma, LP, Cranbury, New Jersey. 

 is with Trace Consulting, Stamford, Connecticut. 

 (Universitätsverlag Wagner, Innsbruck, Austria, 1954).  

y (Interscience Publishers, Inc., New York, 1957), p. 142. 

The authors discuss how applying infrared spectroscopy to microscopic sample preparations extends the value of results generated and provides molecular analysis of physical and chemical interactions.

The Imaging Advantage to FT-IR Spectroscopy

Fourier transform–infrared spectroscopy (FT-IR) offers countless user benefits, including real-time data, multicomponent analysis, and compatibility with automated sampling and microsampling techniques. Types of IR microscope measurement techniques include transmission, reflection, and attenuated total reflectance (ATR). But with each of these, there are pitfalls as well as a number of best practices that will yield superior data. These measurement techniques and their respective analytical processes are described here to help you avoid common mistakes in IR measurement and get the maximum performance from your microscope.

 window materials are often used for Fourier transform–infrared (FT-IR) transmission measurements, allowing for easy collection of infrared spectra. However, each of these requires great care to obtain good-quality spectra. 

A diamond cell incorporates two cell plates that compress the sample. After compression, measurements are conducted on the sample adhered on a single cell plate. If measurements are conducted with both cell plates in place, an interference pattern is superimposed over the sample spectrum. Figure 1 shows analysis of caffeine powder using a diamond cell. The red line shows the spectrum of the sample on a single cell plate; the blue line shows the spectrum measured with the sample sandwiched between two cell plates. 

Figure 1 shows that an interference pattern is superimposed over the sample spectrum measured using two cell plates, such that the baseline exceeds 100%. The interference pattern occurs due to the parallel space between the two diamond disks. Measurement conducted with two cell plates increases the noise near the 2000-cm

  absorption band of the diamond. So, after a sample is crushed between the two cell plates, the sample adhered on a single cell plate is analyzed. Rubber and other elastic samples are exceptions to this rule and are measured while sandwiched between the two cell plates. 

Only a trace sample is required for transmission measurement. Too much sample lowers the baseline and results in a spectrum with saturated peaks. Figure 2 shows an example of this using a caffeine powder sample. The red line shows the spectrum obtained with an appropriate sample volume; the green line is the spectrum for an excess of sample. 

Excessive samples cannot be compressed sufficiently, so measurements must be conducted on thick film. This results in a spectrum that is difficult to analyze due to a lower or inclined baseline and saturated peaks, as shown by the green line in Figure 2. Therefore, the amount of sample must be adjusted to create a thin layer on the diamond cell. 

Background Measurement and Sample Measurement Positions

The positions chosen for background measurement and sample measurement have an influence on the spectrum. Choosing a background measurement position as close as possible to the sample measurement position achieves a good spectrum with approximately 100% baseline. 

Figure 3 shows the spectra for cellulose fiber obtained with background measured in a corner of the diamond cell (BKG1) and with background measured near the sample (BKG2). The amount of transmitted IR light differs between the edges and center of the diamond cell. The amount of IR light is lower if background is measured away from the center of the cell. In this case, sample measurements result in a spectrum with a baseline exceeding 100%. To avoid this, position the sample in the center of the diamond cell and measure the background as close to the sample as possible. If the sample is not in the center of the diamond cell, the background must also be measured near the actual sample position, away from the center of the cell. 

An interference pattern may appear on the baseline if it is measured on a smooth part of the sample surface. Figure 4 shows examples of polypropylene (PP) measurements. The insert shows a photograph of the actual measured positions. If an interference pattern appears on the baseline, it may hide small peaks. 

Do not increase the aperture if the foreign matter (sample) is relatively large. It is more effective to set the aperture to approximately 50 μm and measure the spectra at a variety of positions than to increase the aperture. Constant results at each measurement position indicate that the foreign matter is a single component or a uniform mixture. Sample identification and analysis by spectrum searching are possible on a spectrum not affected by peak saturation or an interference pattern. 

However, differences apparent between the spectra indicate a nonuniform mixture of foreign matter. In this case, differential spectra are determined to obtain a spectrum for each component. 

Figure 5 shows the measurement of foreign matter on the surface of paper. The aperture is set to 30 × 30 μm. The diagram shows a position of strong intensity near 1500 cm

  (S-1) and a position of strong intensity near 1000 cm

  results from a normal component of paper (cellulose). Calculating the difference between the spectra from S-1 to S-2 in Figure 5 gives the spectrum for just the foreign matter. Figure 6 shows the resulting differential spectrum. Spectrum searching of the differential spectrum in Figure 6 indicated the foreign matter to be magnesium carbonate. 

The IR reflection method is extremely simple, requiring little to no sample preparation. However, measurements are limited to certain sample forms. One of the typical samples suited to microscope reflection measurement is a foreign matter adhering to a highly reflective substrate.  

The specular reflection from a thick object adhering to a metallic plate surface is about 5% of the total incident light. When a sample is thick enough, even if some light is reflected from the metal surface, almost zero light will be reflected from the sample surface. The specular reflectance spectrum obtained in Figure 7 displays a distortion in the first derivative spectrum due to a large change in the reflectance (refractive index abnormal diffusion) where strong absorption occurs. However, a spectrum similar to an absorption spectrum can be obtained by applying Kramers-Kronig (K-K) analysis. 

Typically, less light will reach the detector during a reflectance measurement with an IR microscope compared to a transmission instrument. The amount of light is markedly decreased, particularly when analyzing a thick sample in reflection mode, requiring considerable integration to improve the signal-to-noise ratio (S/N). 

Figure 7 shows a measurement example of a glossy coating sheet (about 1 mm thick) on a metallic plate. Measurement was conducted by focusing on the sample surface and setting the aperture to 30 × 30 μm. 

In Figure 7, the peak shows a distortion in the first derivative spectrum as described earlier, unlike that observed in a normal absorption spectrum. Figure 8 shows the result of applying K-K analysis to the spectrum of Figure 7. It is easy to determine from Figure 8 that the coating sheet is polypropylene. 

Figure 9 illustrates the results of a measurement when focusing on the surface of the metallic plate, rather than the sample surface. It is evident that the peaks are saturated in the reflection absorption spectrum due to the thickness of the sample. 

In this way, when the sample is thick, only the specular reflection spectrum can be obtained if the focus is on the sample surface, and qualitative analysis of the sample is possible by applying K-K analysis.  

With samples like powders and fibers, which have rough surfaces, it is difficult to detect only specular reflection components from the sample surface because most of the light is diffused. When sampling is possible, the IR microscope transmission method is suitable, but when the target object is protruding from the surface, the IR microscope attenuated total reflectance (ATR) method is more applicable. 

Analysis of a thin object adhering to a metallic plate surface is illustrated in the schematic diagram of Figure 10. When the sample is thin, incident light is reflected from the surface of the metallic plate and exits. In this case, the spectrum obtained is a reflected absorption spectrum similar to that in transmission measurement when large amounts of light are reflected from the metal surface (blue line in Figure 10). A spectrum of pure specular reflection and reflection absorption is obtained when a small amount of light is reflected from the metal surface. This kind of mixed spectrum creates positive peak distortion (increased transmission) at the higher wavenumber side of the peak position. Conversely, negative distortion (reduced transmission) occurs at the lower wavenumber side. This characteristic is the same as the Christiansen effect seen when using the KBr pellet and Nujol mull methods. (Christiansen effect: A phenomenon in which the amount of transmitted light changes in the vicinity of the peak position due to the relationship between the refractive index and the reflective index. This often occurs with light-scattering samples.) 

Figure 11 shows a measurement example using a thin sample. Measurement was conducted using a 20 × 20 μm aperture setting. Measurements 1 (red line) and 2 (blue line) show results obtained at different positions. 

In measurement 1, a normal absorption spectrum is observed, but in measurement 2, peak distortion is seen in the vicinity of 1500 cm

 . If this type of spectrum is obtained, it is necessary to search for a position that returns a spectrum similar to measurement 1 since the peak position and absorption intensity are different from the proper values. If the target object is extremely small and it is not possible to change the measurement position, use the IR microscope transmission or IR microscope ATR method to obtain an undistorted spectrum. 

The background and sample measurement positions have an influence on the spectrum in the IR microscope reflection method. Figures 12 and 13 show a sample analysis of an object adhering to a metal plate surface. Figure 13 shows the spectra obtained by background measurement at the two positions (BKG 1 and BKG 2), with the sample position fixed on the foreign matter on the plate. 

In Figure 13, peaks are not as distinguishable in BKG 2, a minute quantity of material identical to the foreign matter is stuck to the plate, and the background peaks are offset by the peak of the foreign matter. This illustrates the importance of carefully selecting the background and sample measurement positions in reflection measurements, as in transmission measurements. 

A spectrum search conducted for the red line spectrum in Figure 13 (BKG 1) revealed the substance to be sodium lactate. 

The author discusses how researchers can gain the most accurate and useful data from their equipment by employing careful sampling and microscope techniques and following a few common best practices.