Ponderings from an Analytical Spectroscopist
Neil Everall, Company Research Associate
Intertek-MSG
As an industrial spectroscopist faced with the problem of characterizing a very wide range of materials, from drug distributions in biomedical devices to moving polymer films on production lines, I’ve always been attracted by the versatility of Raman spectroscopy. For example, few other techniques can analyse the chemical and physical properties of materials with either submicron resolution under a microscope, or in-situ in a reactor sited 100’s of metres away from the spectrometer. When I first started my career in industry, Raman was all too often a technique of last resort; it was not widely known, hence we only ever saw the problems that every other technique had already failed to solve! An exaggeration, perhaps, but not much of one. Since those days, advances in technology such as FT-Raman, CCD detectors, volume phase holographic optics and solid state lasers have led to instruments that are smaller, easier to use, faster and much more sensitive than their predecessors. A perusal of the scientific literature, or attendance at meetings such as FACSS or Pittcon, confirms that Raman spectroscopy is being more and more widely applied than ever, and the number of manufacturers offering Raman instruments of varying degrees of sophistication seems to be growing incredibly quickly. From a technical point of view there has never been a better time to be a Raman spectroscopist.
At the moment there are two recent developments which I find very exciting, both of which have the capability to make a major impact on the applications of Raman spectroscopy. The first is related to Raman microscopy; for decades its been known that Raman microscopy offers an order of magnitude better spatial resolution than infrared microscopy, but if one wishes to build up a detailed image of a sample with high spatial resolution over a large area, then this is a very time consuming process. One either has to map the sample point by point, or globally illuminate a large area and build up the hyperspectral data set by scanning the spectral axis wavelength by wavelength. Both approaches have their problems, and ultimately are very time consuming if one needs a full spectrum at each pixel in a large image. Fortunately, instruments have recently appeared which can obtain useable Raman spectra in tens of milliseconds; this means that large maps can now be assembled very quickly, which makes mapping samples such as whole pharmaceutical tablets feasible in hours rather than days.
I’m intrigued by the possibilities; there are obviously some limitations, such as the need for flat samples (there is no point in getting a Raman spectrum in ten milliseconds if it takes three seconds to autofocus on every point), but I’m sure this technology will challenge the current dominance of global near-IR or mid-IR imaging for certain applications. The main generic question which occurs to me at the moment is “will these fast mapping instruments be more prone to fluorescence?” If each sample point is illuminated for less than one second, will fluorescence which is bleached by a few seconds exposure in a normal Raman microscope suddenly re-emerge when each sample point only sees a fleeting burst of irradiation? The second key question is “just how flat does the sample have to be?” As ever, these issues are bound to be sample and application dependent.
The second area which fascinates me springs from the question “if I shine a laser beam on a turbid or opaque sample, which regions of the sample am I actually analysing?” This is the field of Raman photon migration, which has been explored in the temporal and spatial domains, originally by Pavel Matousek and colleagues from the Rutherford Appleton Laboratory, Oxford, UK, and subsequently by a number of researchers including Mike Morris (University of Michigan, US) and Nick Stone (Royal Gloucester Hospital, UK). The basic premise is that if one focuses a laser beam on one part of an opaque sample, and measures the intensity of the Raman signal as a function of distance from the laser focus, it turns out that one can non-destructively depth-profile the sample. This is because, on average, the further the distance from the laser focus a Raman photon emerges, the deeper it was generated. This is the basis of spatially offset Raman spectroscopy, or SORS. It offers a simple way of removing the signals from opaque surface layers while obtaining spectra of subsurface material. There is a corresponding statement regarding the temporal distribution of emitted photons after short pulse laser excitation, but this is an extremely demanding experiment compared with SORS.
These effects are interesting from a fundamental point of view (it turns out that photons can have convoluted flight paths of several centimetres total length in opaque samples), and also because of the widespread practical applications. For example, it is now possible to obtain in-situ spectra of tissues such as bone while looking through thick overlayers of skin and fat; one can analyse samples inside opaque bottles to check for the presence of explosives or hazardous materials (raising the prospect of fast screening at airports and the like); tablets can be analysed inside intact blister packs or bottles to check for counterfeiting or adulteration; and it is possible to obtain spectra of calcified tissue inside thick (cm) layers of muscle. It’s even been shown that excellent spectra of thick opaque samples can be obtained in a transmission rather than backscattering configuration. This is surprising; most people would not expect much Raman light to pass completely through a white tablet. Transmission Raman should offer real benefits for quantitative spectroscopy, because a large volume of material is sampled compared with normal Raman backscattering. All of these effects become apparent once one asks the simple question “how far can a photon move within a nominally opaque sample?” The answer is “much further than you think!” I really like the fact that such simple experiments, which yield quite counterintuitive results at first sight, are now being exploited in real world applications. For those who would like to find out more, I highly recommend Matousek’s review article1.
1 “Deep non-invasive Raman spectroscopy of living tissue
and powders”
Pavel Matousek, Chem. Soc. Rev., 2007, 36, 1292
