Out of the Lab, Into the Field
Raman spectroscopy has undergone a revolution during the last seven to 10 years, largely as a result of massive investment in the telecommunications industry. Improvements have come through what Eric Bergles, VP of sales and marketing at Bayspec (San Jose, CA) calls a “perfect storm” of price reductions, hyper-competition, miniaturization, and reliability. What were previously very large benchtop instruments requiring a Ph.D. operator, liquid nitrogen cooling, and price tags approaching half a million dollars, have evolved into portable and even handheld instruments suitable for both laboratory and field operations.
“The catchphrase here is ‘out of the lab,’” Mr. Bergles says. “And that has been the key to applying Raman to new markets and applications, which are cropping up everywhere.”
Raman is effective for noncontact, nondestructive chemical analysis. Its attraction has been its applicability to many sample types with no preparation required.
One knock against Raman, however, is that many materials exhibit lightinduced fluorescence when common Raman excitation wavelengths of 532 nm or 785 nm excitation wavelengths are used. Fluorescence interference may be tens or hundreds of times higher than the useful Raman signal.
But at longer-wavelength 1064 nm excitation, Raman overcomes interfering fluorescence, enabling the practical application of Raman in situations where it could not previously be used.
Bayspec is targeting instruments employing the new wavelength to food safety; homeland security; forensics; law enforcement; and the chemical, biochemical, pharmaceutical, and biofuels markets. The latter represents an interesting example of Raman’s “new” analytical capabilities.
Algae are one very promising source of biofuels, because lipids make up about 60 percent of their body mass. These fats are easily converted to diesel fuel. During process development, investigators are eager to screen algae culture conditions for their ability to promote lipid biosynthesis in the organisms. Wet chemical techniques for this purpose are time-consuming, whereas fluorescence completely swamps the short-wavelength Raman effect.
By overcoming fluorescence effects, 1064 nm excitation easily unravels the algae’s lipid fingerprint, quantifying fuel- rich cultures within a few minutes. Whereas 785 nm excitation produces only a large fluorescence shoulder, analysis at 1064 nm clearly shows multiple lipid components with no baseline correction necessary.
Over the next decade, Mr. Bergles sees Raman moving into food inspection applications, among others. With up to 40 percent of our produce coming from overseas, monitoring fruits and vegetables for pesticide residues through traditional testing could hold up shipments for days or weeks. “With Raman, you can do spot checks and know instantly if the produce is contaminated.”
Perhaps no industry has been as slow to take analysis “out of the lab” as highly protocol- and regulation-driven pharmaceutical manufacturing. Seven years into the U.S. Food and Drug Administration’s PAT (Process Analytic Technology) initiative, samples from production suites are still walked over to a wet lab, where their analysis can take hours, sometimes days.
For small-molecule drug making, Raman is used to image the distribution of ingredients in pills and to confirm the identity of raw materials at loading docks, but its application is nowhere as diverse or deep as in broader industry. And despite having several applications where Raman would be ideal, biotech manufacturing is still farther behind than smallmolecule drug manufacturing.
A 2011 report from VTT Technical Research Centre of Finland notes that Raman has not been applied to bioprocessing despite its being standard in other process industries.
One downside of Raman as an at-line or in-line bioprocess “laboratory,” says Lee Smith, Ph.D., president of Process Instruments (Salt Lake City, UT), is the presence of highly-colored or fluorescent materials. “When present, they will swamp the weak Raman signal,” he says. A corollary is that Raman is not particularly sensitive below part-per-million concentrations.
“Raman is very good at distinguishing slight structural changes,” notes Maryann Cuellar, an applications scientist at Kaiser Optical Systems (Ann Arbor, MI). Kaiser has pushed Raman for process applications for 20 years, but according to Ms. Cuellar, small-molecule drug development and manufacture have been slow to adopt the technique, and biologicals have been even slower.
She cites reluctance on the part of biotech precisely when other industries are becoming interested to restrictions of 21 CFR Part 11 (electronic signatures), which puts restrictions on software.
In a bioreactor, Raman is capable of detecting analytes currently measured by sensors or off-line analysis (e.g., HPLC and LC-MS). Examples are glucose, lactate, glutamine, glutamate, cell density, osmolality, ammonium, and dissolved gases. Unlike offline measurements, Raman provides numbers in real time.
A group at Biogen Idec (Weston, MA) recently reported the use of real-time Raman spectroscopy for measuring growth and metabolic profiles of mammalian cell cultures, as well as levels of cell metabolites. Currently, cell analysis is conducted offline by removing an aliquot, staining with fluorescent dyes, and sorting cells with a flow cytometer or imaging under a microscope. Bristol Myers Squibb and Johnson & Johnson have also presented data on Raman for bioprocess analytics.
Raman and near-infrared (NIR) are complementary techniques. NIR is based on absorption and Raman on scattering, and their spectra appear as mirror images. Both techniques identify and quantify chemical structures. And unlike mid-infrared, both methods are unaffected by water. Without water interference, there is no need for spectral subtraction.
“We will need a paradigm shift for bioprocessors to adopt Raman fully,” says Ms. Cuellar. She cites as hurdles a lack of trained spectroscopists in the biotech world and a lack of such training among process engineers and biologists. “There is also a shortage of trained chemometricians capable of extracting information from Raman data.” Chemometrics is a technique that applies concentration models to complex mixtures of chemicals. “In a process setting, Raman is about more than just acquiring spectra.”