While the Raman effect has been in college curricula for decades, for much of that time, it has been considered a research tool. The early (c. 1970) models were driven by quite powerful LASERs and had a nasty habit of burning samples (and eyeballs, if proper protection was not used). It was known for quite some time that Raman produces information complimentary to that obtained by mid-range infrared. [Anyone having to perform group analyses to deduce whether a bond is IR-active, Raman-active, or both can appreciate the richness of Raman spectra. Such analyses were the “joy” (read: “bane”) of graduate students for decades.]
Of course, strictly speaking, Raman is not spectroscopy, per se, but a scattering phenomenon. A monochromatic beam impinges on a sample and the light scatters. If you can imagine a lottery, where one ticket in millions pays off, you can have some idea of Raman scattering. In a lottery, the vast majority of tickets have the same value immediately after the winning numbers are drawn (in this example, zero). However, one or two “lucky” tickets are winners and are, by definition, different from the rest.
Figure 1. Energy levels giving rise to Raman spectra.In the same way, the overwhelming number of photons striking a sample are simply “elastically” scattered (the Rayleigh scattering) and emerge from a sample cell with the same energies (frequencies) with which they entered. The “winning tickets” or photons that were “inelastically” scattered and, depending upon which bond they “tripped over,” have either higher or lower energies than the incident radiation (Stokes, lower energies or anti-Stokes, higher energies. See Figure 1). This resultant “spectrum” is a fingerprint of the material and many of its physical properties.
By physical properties, I refer to poly-morphism, due to either the presence or absence of solvents (see Figures 2 and 3) or other crystalline forms. Since Raman only “sees” molecules with a center of symmetry, water is all but invisible (which is why it is detected through its effects on the crystalline structure of a molecule). This make Raman a perfect tool for aqueous samples (where IR is totally swamped or NIR, which is pathlength limited). Also, since it is a scattering, not transmission, mechanism, there is no need for an opposing detector (cuvette) or dipping tube with reflecting mirror.
Figure 2. Raman spectra of Citric Acid polymorphs. Red = anhydrous; blue = monohydrate
Figure 3. Raman spectra of anhydrous dextrose (red) and dextrose monohydrate (blue)These attributes have always been available in bench-top instruments, but, opening containers (for incoming raw materials, for instance), labeling and bringing samples to the lab, placing them into proper containers, and running them is only marginally easier than “traditional” raw material analyses.
What has been attempted for years was to design and build a portable (for ordinary technicians, not just for weightlifters) Raman unit for use in the field. The “field” can be anything from the warehouse to an importation center (Customs House) to a local hospital or pharmacy; in other words, anywhere that isn’t the traditional Analytical Lab. [Although the latest models are quite capable of lab level results.]
Some of the problems, aside from the safety factor in walking around with a powerful LASER on your hip, were largely in making that LASER give the same results, time after time. We have moved from LASERS to LASER diodes, to Light Emitting Diodes (LEDs) and Super-Luminescent LEDs (SLEDs) as light sources. While the spectra generated were always recognized as “Ramanesque,” the reproducibility of the signal and potential interference of Rayleigh-scatted light made the signals less than perfect for pharmaceutical applications. In addition, the quantitative part was somewhat elusive (at best). The FDA had, for quite some time, expressed doubts about the ability to quantify samples and, accordingly, tended to dismiss Raman applications in NDAs and ANDAs.
A number of operational problems have been addressed since the first attempts at “portable” units were introduced, Using input from the communications industry, stable, efficient sources have been introduced for Raman units. [With the collapse of the telecom industry in the 1990s, a number of former telecom suppliers had to either find new customers (uses for their wares) or close their doors.] The problem of fluorescence was solved, initially, by producing light at increasing wavelengths; as the wavelength moves from visible to near-infrared, the corresponding fluorescence also diminishes. That, as the saying goes, was the good news. The bad news? The Raman signal diminishes by the fourth power of the wavelength used to generate the effect.
Another problem is the Raleigh or elastically scattered light. It remains at the incident wavelength and is orders of times stronger than the Raman radiation. One solution was to use a “notch filter” or interference filter that allows wavelengths above and below a central (spectral) region of the incident light. [Any interference filter or dichroic filter is an optical filter that reflects one or more spectral bands or lines and transmits others, while maintaining a nearly zero coefficient of absorption for all wavelengths of interest. An interference filter may be high-pass, low-pass, band-pass, or band-rejection; for Raman, band-rejection is used. It consists of multiple thin layers of dielectric material having different refractive indices or metallic layers. Interference filters are wavelength selective by virtue of the interference effects that take place between the incident and reflected waves at the thin-film boundaries. The important characteristic of the filter is the form of the leaving signal. It is considered that the best form is a rectangle.]
With better light sources and potentially game-saving solutions to fluorescence and scattered light, a number of modern, handheld units have been produced. A relatively recent review of several units was published as a reference. While not comprehensive, it is a start. (http://nortonsafe.search.ask.com/web?q=handheld+Raman&o=apn10506&prt=cr) Several points were brought up and need to be addressed before handheld units become ubiquitous.
- Cost. While the cost per unit has come down in recent years, most are well above $10,000. There has been a push in the last few years to develop UV/Vis/NIR units, capable of being attached to smartphones and able to produce credible spectra… for hundreds of dollars, not thousands. If a way to produce Raman handhelds in the $1,000 range (a ten-fold decrease in price), the use will soar, not just in Pharma, but in every industry.
- Reproducibility. The ability of one company to demonstrate that unit to unit, their instruments can produce identical spectra. This seems fairly good for the majority of companies, but the intensities on many units may have trouble reproducing, hence, the reluctance of agencies (FDA, EMA) to embrace quantitative applications of handheld units.
- Quantification. While not a “game-breaker,” the questions about quantitation abilities are tending to keep Raman as a qualitative tool (certainly not a bad thing) and limiting its potential. Much of this problem is likely traced to the fact that many “warehouse” users are not technicians and, as such, may not grasp the importance of sample presentation, among other things.
- Method Transfer. Assuming the quant and qual questions may be answered, there is still far more work being done on NIR instruments as far as method transfer from manufacturer to manufacturer is concerned. I am sure the time will come, but it is not here yet.
- See #1. All goes back to cost. It is a vicious cycle, of course. The cost will come down as many units are sold, but to sell any units, the cost has to be a lot less. I am not an instrument manufacturer, so I don’t have all the answers. One might include the words, “loss leader” for some larger, multi-technology instrument company, where they sell below cost until the cost comes down to where a profit may be realized. Just saying.
Summary
I personally believe that Raman spectroscopy is very important, both as a research tool and a process/quality aid. Its use as a handheld unit will allow analysts to range far and wide and bring the lab to the samples, much as mobile units were used in-house at various location.
There is no question about the quality of spectra and ability of the Raman technology to glean incredible amounts of information from field and production samples, but the current units need to be a tiny bit more reproducible and a lot less expensive. Then we have one of the strongest tools in our analytical toolbox. They are good and will get better, I believe.
Author Biography
Emil W. Ciurczak has advanced degrees in Analytical and Physical Chemistry from Rutgers and Seton Hall Universities. He has been working in the pharmaceutical industry since 1970, performing analytical method development and method improvements, as well as designing and modifying analytical devices. Some of his publications are listed in the HYPERLINK "Text_Books_and_Chapters.html "Text Books & Chapters section. He is the winner of the EAS (2004) Award for Contributions to NIR. His column for Pharmaceutical Manufacturing, Therapeutic Dose, won the ASBPE 2007 Gold Award for "Regular Column, Contributed." In addition to pharmaceutical companies, Emil also worked for Union Carbide (polymers) and Henkel Corp. (surfactants). In all his positions, he has continually introduced new technologies. He was a pioneer for near-infrared (NIR) for the pharmaceutical industry, first introducing it at Sandoz in 1983 for raw material qualification. Some of the properties of pharmaceutical materials (using NIR) reported by (published and presented) Emil include polymorphism, purity of optical isomers, mean particle size, dosage form assay, verification of clinical supplies verification, counterfeit detection, and identification/qualification (he the was "beta site" for pioneer software from Technicon) of materials and dosage forms. He applied NIR to Chromatographic and dissolution testing, as well as for solid dosage form analyses. Some of the instrument companies with whom Emil has worked include Metrohm NIRSystems, Technicon (Bran+Leubbe), Buchi, Brimrose, Control Development, Infrared Fiber Systems, LT Industries, Spectral Dimensions (now Malverne), and Bruker. He is named on over a dozen patents for instruments and software, largely for process control and Chemometrics. Emil has been an adjunct professor since 1979, teaching (college-level)courses in General Chemistry, Analytical Chemistry, Instrumental Analysis, Instrument Troubleshooting, Physics, Thermodynamics, and has conducted (award winning) student research. Emil was on the USFDA-PAT subcommittee (Validation) in 2002-3, helping CDER gather information for the Guidance it issued. This has helped in formulating the QbD/PAT, Design of Experiment, and instrumentation (NIR, Raman) courses (Pharma and BioPharma) for the HYPERLINK "http://www.cfpa.com/"Center for Professional Advancement, HYPERLINK "http://www.informaglobalevents.com"PTi, and HYPERLINK "http://www.informaglobalevents.com/division/%E2%80%8Blife-science"Informa LS