As shown in Table II, ethyl cellulose and hydroxypropyl cellulose also correspond to the longest required measurement times,
284 s and 390 s, respectively. Further inspection of the spectral data (not shown) revealed subtle features in the unknown
measurements that were not present in the reference measurements. To determine whether the polyethylene containers could be
responsible, an auto-mode spectrum for each of these two materials was acquired through borosilicate glass vials. These spectra
were consistent with the reference measurements and did not contain any extra features. In addition, a spectrum of polyethylene
was acquired by folding an empty polyethylene bag over on itself several times to produce a sample of sufficient thickness
for measurement. The resulting spectrum (not shown) contained bands in the spectral regions where the extra peaks in the unknown
were found. Thus, despite the ability of Raman to sample through packaging, the extremely weak signal and long measurement
times for these two cellulose materials resulted in conditions favorable to allow the plastic to subtly interfere with the
measurement. It is likely that measurements could be successfully made through glass, but this was not attempted because the
required measurement time may not be considered practical for routine field use.
Examination of the off-diagonal elements in Figure 6 confirms the excellent selectivity of the technology as evidenced by
the overwhelming majority of off-diagonal elements with p < 10–15. The only area in the table where there is a lack of acceptable selectivity is for the alkali metal stearate materials. Stearic
acid is differentiable from both calcium and magnesium stearate (p < 0.01); however, calcium stearate and magnesium stearate cannot be readily differentiated from one another. Whereas many
materials differing only in their cation can be readily differentiated with the handheld Raman system (e.g., in the study,
sodium and potassium bicarbonate, calcium, and zinc sulfate), calcium stearate and magnesium stearate are simply too similar
from a spectroscopic standpoint. This is likely a result of the change in cation not having an appreciable long-range influence
on the relatively large anion (stearate) from which the Raman signal is actually generated.
Conclusion
Handheld Raman spectroscopy is an excellent alternative to traditional incoming raw-material inspection by high-pressure liquid
chromatography, wet chemical methods, and NIR and mid-IR spectroscopy. The technology has excellent specificity, which, coupled
with intelligent on-board algorithms, reduce the time and effort required to develop and validate methods. Furthermore, methods
can be successfully loaded onto different handheld Raman instruments to produce consistent data and material identification
on the multiple instruments, without loading additional spectra or performing other customization.
In addition to its analytical characteristics, today's handheld Raman solutions are environmentally robust and can be used
by expert spectroscopists as well as operations-based personnel. This is in contrast to Raman instruments of the past, which
were bulky, slow, expensive, and delicate. Based on the study presented in this article of common pharmaceutical materials,
the handheld Raman spectrometer offers an attractive option for 100% inspection of most incoming raw material in pharmaceutical
manufacturing facilities.
Acknowledgment
Dr. Wayne Jalenak is gratefully acknowledged for useful discussions and for preparation of the glycerin samples presented
in the introduction.
Robert L. Green* is a research scientist and Christopher D. Brown is a director of system analytics and applications, both at Ahura Scientific, Inc., 46 Jonspin Road, Wilmington, MA 01887,
tel. 978.657.5555, fax 978.657.5921, bgreen@ahurascientific.com
*To whom all correspondence should be addressed.
Submitted: July 26, 2007. Accepted: Sept. 13, 2007.
References
1. H.T. Rasmussen et al., "Validation of HPLC Methods in Pharmaceutical Analysis," in Handbook of HPLC in Pharmaceutical Analyses, S. Ahuja and M. Dong, Eds. (Elsevier, San Diego, United States, 2005), pp. 192–216.
2. M.A. Dempster et al.,"A Near-Infrared Reflectance Analysis Method for the Noninvasive Edentification of Film-Coated and
Non–Film Coated, Blister Packed Tablets," Anal. Chim. Acta. 310, 43–51 (1995).
3. M. Blanco et al., "Near-Infrared Spectroscopy in the Pharmaceutical Industry," Analyst 123, 135R–150R (1998).
4. P.J. Gemperline, L.D. Webber, and F.O. Cox, "Raw Materials Testing Using Soft Independent Modeling Class Analogy Analysis
of Near-Infrared Reflectance Spectra," Anal. Chem. 61 (2), 138–144, (1989).
|
