3. Vibrational Microscopy:
A Personal Guide to Techniques and Applications.
Pfizer Central Research,
The term vibrational microscopy is one that is becoming more frequently used to describe collectively the techniques of FT-IR microscopy, NIR microscopy and Raman microscopy. All allow vibrational spectra (i.e. IR or Raman) to be obtained from small samples via adapted light microscopes. This paper is an introduction to the practical aspects of vibrational microscopy, the types of experiments that are currently possible, and some interesting applications of these versatile methods. For more information, reviews,[1, 2, 3] reference books,[4, 5] and other journals should be consulted. It should be pointed out that the nomenclature used to describe these techniques is not globally consistent. European publications tend to use the terms FT-IR microscopy and Raman microscopy. This is consistent with the nomenclature used by the Royal Microscopical Society (e.g. Light microscopy, Scanning electron microscopy etc.). American publications use the terms IR microspectroscopy and Raman microspectroscopy. The term microspectroscopy is also appropriate for microsampling techniques which include beam condensers, diamond anvil cells, and the new generation of single reflection ATR accessories, as well as being used to describe microscope based systems. My personal belief is that the European nomenclature is scientifically specific and correct. However, with the dominance of the US publications and market forces it is unlikely that there will be a unification of terminology in the foreseeable future.
Historically, it was the Raman microprobe  which first made an impact on small sample analysis. However, this type of system was typically only found in multi-national companies and a few academic institutions. The applications afforded by FT-IR microscopy, and introduced to the scientific community in the early 1980’s, proved to be much more applicable and accessible to many more analytical laboratories. Raman microscopy is currently going through a renaissance due to the emergence of bench top systems using CCD (charge coupled device) detectors and notch filters (rather than double or triple monochromators) for laser rejection. NIR microscopy is an emerging technique which has effectively extended the spectral range of the FT-IR microscope into the near IR region. Its applications have yet to be established. It should be stressed that vibrational microscopy is only one of a number of spectroscopic and microscopical tools the analyst can call on to solve problems. There are many forms of microscopy that can give information on the chemical composition and/or sample structure. Many of these techniques are included in a review paper by Peter Cooke on chemical microscopy and this is a good reference point for identifying modern microscopy techniques and their applications.
The bulk of this article concentrates on how to carry out vibrational spectroscopy and its current applications. With the wide scope of information, it has been divided into the following sections.
Practical aspects of vibrational microscopy
i. Sample preparation for FT-IR microscopy
ii. Sample preparation for Raman microscopy
Recent applications and technological advances
i. Instrumentation and techniques
ii. Areas of Application
Practical aspects of vibrational microscopy
When or how should an FT-IR or Raman microscope be used ? This question will have a wide variety of answers depending on what information is required. Vibrational microscopy is used in its most simple form as an expensive beam condenser coupled with a high performance viewing system. This allows the analyst to survey the sample and position the area of interest in the IR or laser beam. At the other extreme it may be used to determine the number and physical shape and size of the different phases present. By observing samples under white light and through crossed (plane) polarisers, areas of birefringence will be observed enabling identification of crystalline materials to be made. If the system is fitted with photomicrography or image capture equipment, a permanent white light image of the sample can be obtained.
With respect to size, transmission FT-IR microscopy measurements can be made on samples having XY dimensions of ca. 10-250 m m with a thickness of <30 m m. Typically samples have dimensions of several 10’s of microns. Recent developments using synchrotron sources enable sub 10 m m samples to be studied . Raman microscopy systems allow significantly smaller samples to be studied routinely. Depending on whether a dispersive or FT instrument is used the sample size is typically 1-30 m m.
Before performing any vibrational microscopy, I would recommend observing the sample using a low power stereomicroscope. This has two benefits. Firstly, it allows the whole sample to be surveyed quickly, and as these microscopes are binocular instruments, a full perception of sample depth is provided. This simple sample observation may give many clues as to the samples identity, or to what has happened to the sample prior to analysis. From my own experience dark coloured features that appear to be on the surface of a sample when viewed by the naked eye, may in fact be close to the surface, but covered by clear or less coloured matrix. Secondly, if required, the stereomicroscope can be used as an aid in microsampling preparation. Its low power magnification (10-60X) and large working distance* make, for example, the extraction of a specific particle or fibre from a solid or liquid matrix with a fine needle considerably easier than performing the same operation whilst viewing the sample by the naked eye alone.
(* working distance = distance between objective lens and sample stage)
i. Sample preparation for FT-IR microscopy
Where possible, I prefer to obtain spectra in transmission mode as these tend to produce significantly better spectra than those made in reflection mode. This generally requires some form of sample preparation to maximise the sample area and minimise its depth. ( My general rule of thumb is that if the particle can be observed by the naked eye it is too big to be used “as is” for transmission FT-IR microscopy work). The sample will often require some form of pressing, squashing or rolling before it is transferred to an optically clear window material suitable for FT-IR microscopy (e.g. NaCl, BaF2 etc.). An elegant and effective preparation method is to press the sample between two diamond windows. This method offers three advantages which are:
1. the sample is contained by the device avoiding accidental sample loss during flattening.
2. the procedure can be observed under a stereomicroscope. Pressure is only applied until the sample forms a glass and flows to form an appropriately thin sample.
3. the diamond can be used as the window material, thus removing the need to place the prepared sample on another substrate prior to FT-IR microscopy.
In some cases the sample will have to be studied without preparation. This may be because the sample could change during preparation (e.g. undergo a polymorphic transformation), or is required in its intact state for further testing or archiving. Where these samples are too thick or opaque for transmission measurements, the objective Cassegrainian lens can be used in reflectance mode. It can be considered to be a micro diffuse reflectance accessory; one half of the lens delivering the IR beam to the sample and the other half collecting the resulting reflected energy. This (transflection) approach provides ca. 50% of the incident energy available in transmission experiments, so in practical terms reflectance measurements require extended acquisition times in order to obtain the same quality spectra that are obtained in transmission mode.
Figure 1. ATR-FTIR microscopy in action
There are two other approaches to reflectance FT-IR microscopy. These are through the use of attenuated total reflectance (ATR) – See Figure 1, and grazing angle objectives. ATR FT-IR microscopy uses the same principles as traditional ATR methods and allows spectra of microscopic areas to be obtained through direct contact with the ATR objective – See Figure 2. This method is so easy to use it can also be used to obtain information from sample surfaces that would normally be obtained using macro ATR accessories. (N.B. this approach has been modified in the design of the latest single reflection ATR units that are proving extremely useful in providing rapid analysis of bulk solids, liquids, and pastes). Grazing angle accessories have been used for many years to study thin coatings on substrates, and thin film lubricants. The availability of grazing angle objectives means that these types of studies can now be performed routinely at a microscopic scale.
Figure 2. ATR-FTIR microscopy sample in contact with ATR objective
Another approach to obtaining spectral information from thick or opaque samples is to use NIR microscopy. NIR bands originate from overtone and combination bands of IR fundamental bands and as a consequence are significantly weaker than those in the mid IR region. This is advantageous in NIR microscopy as much thicker samples
(>1000 m m ) can be studied without exceeding the dynamic range of the detector. This also means that minimal sample preparation is required prior to performing NIR microscopy. This approach was first suggested in the late 1980’s, but has only recently become a commercial reality.
ii. Sample preparation for Raman microscopy
Raman spectroscopy is a complementary technique to IR spectroscopy in terms of the spectral information each provide. In vibrational microscopy there is also a complementary nature to sample analysis. In the Raman experiment, the Raman shifted wavelength are backscattered 180° to the incident laser radiation. For this reason virtually no sample preparation is required for many samples. Sample thickness and opaqueness for Raman measurements are irrelevant, and those samples which are difficult to study by FT-IR microscopy pose few problems for the Raman microscope. Similarly, samples containing water or in contact with glass can be measured directly by Raman microscopy. Generally all that is required is that the working distance of the microscope can accommodate the sample.
The major problem associated with any Raman measurement, including microscopy methods, is sample fluorescence. Fluorescence can often be avoided by using FT-Raman systems using 1064 or ca.780 nm lasers to irradiate the sample. Unfortunately, sensitivity at these wavelengths is poor. Sensitivity is improved by using an excitation wavelength of 633 or 514 nm, but here fluorescence may be the dominant spectral feature. A new approach to avoiding fluorescence is to use a deep UV laser, but by this method the sample may be damaged due to the much higher energy of this excitation source. In my experience, a suitable laser wavelength with which to obtain Raman spectra can be found for the majority of samples submitted for analysis.
Recent applications and technological advances
i. Instrumentation and techniques
The most current and significant area of growth in vibrational microscopy is chemical mapping and imaging. Although the concept is not new, the instrumentation required to provide reliable IR or Raman chemical images has only become available over the last few years. This technology now allows multi-component systems to be mapped, and the location, shape and size, of each material present to be visualised through the chemical images that are produced. These images are based on spectral features that are unique to each of the materials present and provide a valuable source of chemical information that has previously been unattainable. Specific instrumentation and applications are detailed elsewhere in this journal [10, 11] and other review articles.[4, 12] All these mapping/imaging techniques produce chemical images with a spatial resolutions of ca. 1-20 m m. A development of SNOM (Scanning Near-field Optical Microscopy) has reduced this spatial resolution below the diffraction limit and provides a spatial resolution of between 20 and 150 nm from which Raman spectra can be obtained. [13, 14] The technique can be described, in very simple terms, as the linking of the atomic force microscope (AFM) to a spectrometer via a fibre optic. SNOM is an emerging technique and has many novel applications including the study of receptor / ligand binding in biological systems, thin films (e.g. Langmuir-Blodgett monolayers), and stress measurement in silicon. The technique of photothermal FT-IR spectroscopy has recently been announced which allows discrete thermal and IR data to be obtained at a similar sub micron spatial resolution.
Several classical light microscopy techniques have been successfully adapted for use in vibrational microscopy. A technique which is becoming increasingly more used is confocal Raman microscopy. This exploits a feature found in some optical microscopes. Using an aperture at one of the microscopes focal planes, it is possible to study non-invasively a layer within a sample rather than just at its surface. This allows a sample to be depth profiled, and if a XY stage is available, provides an opportunity to image a sample volume rather than just a surface.
Crystallinity and molecular orientations in individual particles and fibres can be probed by using polarised FT-IR microscopy. Here an infrared polariser is placed in the IR beam and spectra acquired from the sample placed alternatively perpendicular and parallel to the polarised beam. Differences in the spectra can be used to determine specific orientations and interactions of functional groups within the sample. In the case of fibres, dichroic ratios (Iparallel/Iperpendicular) can be used in the non-destructive prediction of their physical properties .
FT-IR and Raman thermomicroscopy use a heated / cooled stage to enable samples to be studied at non-ambient temperatures. The main applications of vibrational thermomicroscopy are to study spectroscopically (and visually, though not necessarily mutual) the changes in polymorphism and crystallinity of single sample particles/crystals. In these experiments the transformation from one form or phase to another is temperature dependent. Another approach is to use a modified differential scanning calorimeter to heat and cool the sample. This combination of techniques allows structural information to be correlated to the thermogram obtained from the same sample.
ii. Areas of Application
The ability to perform non-destructive or non-invasive analysis on small and often unique samples by vibrational microscopy has been utilised successfully by forensic scientists. There are many examples in the public domain of how these techniques have been used to identify microscopic particles, crystals, paint chips, and fibres associated with accidents, incidents, and criminal activity. Constant improvements in instrumentation now allow not only fibres identities to be determined, but also the dye types used to colour them . This is potentially very beneficial in understanding and dating historical samples. Similarly, Raman microscopy is being used in art conservation  to determine which pigments are deteriorating and those which have been added to the piece at a later date. This thinking can be extended to identifying forgeries, by comparing the date of the painting with those of the pigment introductions to the art world. Any anomalies between dates would suggest the history and/or origin of the painting should be investigated.
An exciting application has used Raman imaging techniques to study fingerprints for traces of explosives such as TNT and Semtex.  The imaging identifies which components are present, and by making confocal measurements an estimate of the amount of each compound present across the image can be made.
Vibrational microscopy has many applications in biomedical studies (extensively reviewed by Victor Kalinsky ), including an often quoted example of the imaging and identification of silicone in breast tissue biopsies following the failure of implants. FT-IR can be used to differentiate between the a -helix and b -sheet secondary structures in proteins. This has been used successfully to show that the anticancer drug adamantyl maleimide ( AMI) produces a conformational change in both the membrane and intracellular proteins of the human gastric carcinoma cell line SC-M1 . This gross change in secondary structure has been linked to the cell viability, and can be used to monitor the effect of different drug concentrations on the carcinoma.
Principle component analysis and reflectance FT-IR microscopy data has been used to differentiate between the properties of bacterial colonies . By observing the frequency of the asymmetric phosphate band, it was possible to determine which colonies were Gram positive and which were Gram negative. The advantage of this technique is that it is non destructive, requires no sample preparation, and is better at discriminating between bacteria types than other spectroscopic methods. The use of synchrotron sources has allowed IR spectra to be obtained from single living cells . With spatial resolution of a few microns now possible, the distribution of functional groups of proteins, lipids and nucleic acids in a cell have been mapped. In addition, the changes in lipid and protein distributions during cell division and necrosis have been determined using this method.
Within the pharmaceutical industry, vibrational microscopy has a special role in the analysis of samples in their solid state. This is important as the polymorphic form of the drug and/or excipients is often vital to the performance of the product. For example , a capsule formulation of SCH 48461 showed a slowing of dissolution with respect to time. Using ATR FT-IR microscopy this was shown to be due to a conversion in the formulation of the drug from an amorphous state to crystalline form; the latter having a lower solubility in the water. Similarly, Raman chemical imaging has been used to determine the size, form, and distribution of the drug and excipients in tablet formulations. Figure 1 shows the chemical images from two batches of a tablet formulation which have good and poor dissolution profiles. These images show that the particle size and distribution of the drug and excipient A are different in the two batches. With this information it has been possible to optimise formulation manufacture to produce tablets with good dissolution properties .
Figure 3. Raman chemical images from tablets having different Dissolution profiles.
Scale: each pixel is 5 x 5 µm.
KEY: Red = Drug; Blue = Excipient A; Green = Excipient B
Another advantage of chemical imaging is that it can be applied to complex multiphase systems, which by light microscopy are of inherently low contrast. Emulsions are good examples of such systems, and Raman imaging offers a unique method for identifying the components present in aqueous, droplet and structural phases present in these formulations . A review of this type would be incomplete without a few recent examples from the polymer industry. A paper by John Chalmers and colleagues  demonstrates the applications of several of the techniques in the characterisation of industrial materials. The use of a synchrotron as a light source now allows the IR spectra of sub 10 micron thick samples to be obtained, which has applications in studying thin layers in polymer laminates. One example demonstrates how IR imaging shows not only the location of known polymeric material in a sample, but also locates and allows identification of contaminants which were invisible by light microscopy. Imaging is not confined to static systems. Time domain FT-IR microscopy utilising a focal plane array ( FPA) detector has allowed the observation of low molecular weight liquid crystals diffusion into poly( butylmethacrylate) film in real time .
The performance of a polymer is dependant on a number of factors, a number of which can be probed at the microscopically level by vibrational microscopy. FT-IR microscopy in conjunction with scanning electron microscopy (SEM) and light microscopy, has been used to assess the crystallinity and molecular orientation and by implication, the performance of nylon 6,6 vibrational welds . There are many examples of the Raman spectroscopy/microscopy to determine the stress/strain present in a sample. A recent example demonstrates how this microscopical method can be used to measure these properties at localised spots in polymer based composites .
Mention of other applications should be made which include quality and microstructure measurements of CVD (chemical vapour deposition) diamond films used as coating for computer hard discs, analysis of combinatorial chemistry products whilst attached to their solid phase substrates, and the investigative analysis of unknown fibres, particles, crystals and contaminants.
Although, not an exhaustive review of vibrational microscopy applications, this article has given an overview of the type of studies that can be tackled using FT-IR and Raman microscopies. The continued development of new instrumentation and on-going research projects clearly demonstrates that vibrational microscopy has established itself as a valuable tool in solving microscopical problems in analytical laboratories around the world.
- Katon J.E., Micron, 1996, 27(5), 303-3182
- Clark D.A., Encyclopedia of Anal. Science., Academic Press, 1995, 3174 – 3182.
- Treado P.J. and Morris M.D.,Appl. Spectrosc. Rev., 1994, 29(1), 1-38
- Infrared Microspectroscopy Theory and Applications; ed. Messerschmidt R.G. and Hartcock M.A.; Marcel Deker Inc. New York, 1988.
- Practical Guide to Infrared Microspectroscopy; ed.Humecki H.J.; Marcel Dekker Inc. New York; 1995
- Delhaye, M. and Dhamelicourt, P.J.; J. Raman Spectrosc.; 1975, (3), 33-43
- Cooke, P.M.; Anal. Chem., 1998,70(2) 385R-423R
- Reffner J.A.,Carr, G.L, Williams, G.P., Mikrochim.Acta, Suppl., 1997, 14(Progress in Fourier Transform Spectroscopy), 339-341
- Smith M.J., Appl. Spectrosc., 1989, 43(5), 865-873
- Treado P.J.; IJVS in press
- Wright N..,IJVS in press
- Lewis E.N., Levin I.W., J.Microsc.Soc.Am., 1995, 1(1), 35-46.
- Webster S., Batchelder D.N., Smith D.A.; Appl. Phys. Lett.; 1998,72(12) 1478-1480.
- Pohl D.W et al.; Chimia 1997, 51(10), 760-767.
- Grieve M.C., Griffin R.M.E.; Malone R.; Sci. Justice 1988, 38(1), 27-37.
- Clark R.J.H., Gibbs P.J.; Anal. Chem. 1988, 70(3), 99A-104.
- Mercado A.G.; Janni J.; and Gilbert B.; Proc. SPIE-Int. Soc. Opt. Eng.; 1995, 2511, 142-152
- Kalinsky, V.F.; Appl. Spectrosc. Rev.;1996, 31 ( 3) , 193-249.
- Wang, J-J; Chi, ,C-W; Lin, S-Y; Chern Y-T; Anticancer Res.; 1998 17 ( 5A) , 3473-3477.
- Lang, P.L. and Sang, S-C.; Cell. Mol.Biol.; 1998, 44 ( 1) , 231-238.
- Jamin, N. et al.; Proc. Natl. Acad. Sci. USA.; 1998, 95 ( 9) , 4837-4840.
- Markovich, R.J. etal.; J. Pharm. Biomed. Anal.; 1997, 16 ( 4) , 661-673.
- Clark, D.A. and Staps, D.J.; Proc. 3rd Aust. Conf. Vibrational Spectrosc., 1998, 31-33.
- Andrew, J.J. et al.; Appl. Spectrosc.;1998, 52 ( 6) ,790- 796.
- Chalmers, J.M. et al.; Analyst; 1998, 123 ( 4) ,579-586.
- Snively, C.M. and Koenig, J.L.; Macromolecules; 1998, 31 ( 11) , 3753-3755.
- Stevens, S. M.; Annu.Tech.Conf. – Soc. Plast. Eng.; 1997, 55 ( 1) , 1228-1232.
- Arjyal,B., Paipetis, A., Galiotis, C.; Nondestr.Test. Eval.;1996, 12( 6) , 355-366.