2.1

See through identification

In conventional Raman, the excitation laser is focused on or close to the sample surface, as shown in Figure 1(a). When this surface is non-transparent packaging such as a white polyethylene (PE) bottle, the Raman scattering from the bottle surface is strong, and only a small amount of the laser penetrates through the wall to the sample inside. The spectrum 1(a) below of is of a sample of sodium benzoate in the white bottle, where the arrows are pointing to the weak signal of the sodium benzoate, with the predominant signal arising from the Raman scattering of the PE bottle. Using a defocused excitation beam and a diffuse scattering mechanism, the light is covering a larger area (several mm) and penetrates through the bottle wall. In this case the Raman scattering of the sample beneath the surface is measured with less of a contribution from the surface itself. The spectrum in 1(b) has the signal of the sodium benzoate as the predominant peaks, though their overall intensity is not that high, and the bottle surface has a much smaller overall contribution. Using the reflective cavity of the STRaman technology as shown schematically in figure 1(c), the Raman scattering signal as well as the excitation signal are enhanced by being reflected in the cavity and having multiple bounces on the sample, over the larger sample area and to a greater depth. The STRaman technology results in a Raman spectrum with overall greater intensity dominated by the subsurface sodium benzoate peaks, marked with arrows in figure 1(c), with minor contribution from the PE bottle surface. The reflective cavity increases the signal throughout almost 8x.

Figure 1:

Spectra of sodium benzoate in a white polyethylene bottle in the (a) standard Raman configuration; (b) defocused laser excitation covering larger area, deeper volume of sample and (c) with reflective cavity of STRaman technology with much enhanced signal over the larger area and beneath surface. (785 nm laser excitation, 1 sec integration, ~ 300 mW laser power).

This example illustrates the ability for subsurface sampling such that sample can be identified with minimal handling even when the packaging is opaque. This is clearly beneficial in minimizing the need to perturb samples, and identify samples for quality control, and to detect mislabeled and unknown materials as may be encountered by first responders, customs, postal service and shipping companies.

The identification of samples such as coated tablets is important within the safety industry, and by customs and regulators for the detection of controlled substances as well as counterfeits. Tablets are often coated with sugar and with colorants or titanium dioxide to make them more palatable, increase tablet durability, control the dissolution properties, and to distinguish the product with distinctive branding. In order to identify the tablet’s active ingredient (s) it is necessary to break the tablet, remove the coating layer or have a method to measure through the tablet coating. The spectrum in Figure 2c of an Advil® tablet with the coating removed shows the clear distinct Raman spectrum of ibuprofen. Using library search algorithms gives users a rapid means of identifying the sample as ibuprofen. When this enteric-coated tablet was scanned by conventional Raman scattering, the major contribution to the spectrum is the sucrose coating, where the peaks in Figure 2(b) from 600 to 1400 cm^-1 are attributable to the sucrose, while the peaks of the active are absent. The see through scan of the coted table shown in Figure 2(a) has contributions of both the active ibuprofen with small peaks of sucrose also visible, The strong signal of the ibuprofen, marked by arrows has a close resemblance with the bottom spectrum where the coating is removed. Specialized library searching methods using proprietary algorithms have been developed to isolate the sub-surface Raman spectrum from the surface spectrum so that the sample can be identified, without confounding results from the packaging or other surface layer signal.

Figure 2:

Raman spectra of coated tablets. (a) Coated tablet spectrum measured using the STRaman technology; (b) Coated tablet spectrum measured with a conventional confocal Raman configuration; (c) spectrum of the tablet with the coating stripped measured with the conventional confocal configuration.

An additional benefit of this STRaman system is that even colored tablet coatings can be measured without sample surface burning due to the lower power density of the system relative to that of the highly focused signal used in conventional Raman.

2.2

1064-nm laser STRaman for sampling through fluorescing layers

There are many cases in which multiple layers of packaging are used for samples, including street trade of illicit substances. Many raw materials for pharmaceutical and consumer products are supplied in kraft paper sacks of single or multiple layers, sometimes with a plastic lining layer as well. Brown kraft paper, made from chemical pulp, has a strong fluorescence when measured with 785 nm Raman as do many other dark colored materials. The increased penetration achieved with the STRaman technology on a 1064 nm Raman system enables material identification through such challenging surface layers. This technology has been used to identify materials from weak Raman scatterers such as trisodium phosphate to the strongly Raman active calcium carbonate through different multi-layer paper bags. Successful identification has been made based on the Raman spectrum measured through bags with multiple layers of kraft paper, white paper and plastic film.

The 1064 nm Raman spectrum of trisodium phosphate (Na₃PO₄) collected with see through reflective cavity adaptor as measured through a two-ply bag of white and brown kraft paper is shown in Figure 3. Although the spectrum in 3 (a) is dominated by fluorescence and Raman features from the paper layers, the library matching algorithm used is capable of reliably extracting the spectrum of trisodium phosphate from the packing layers and positively identifying it.

Figure 3:

1064 nm STRam identification of trisodium phosphate (Na₃PO₄) in bilayer bag of white and brown kraft paper. (a) Spectrum measured through the bilayer bag using the STRaman technology; (b) spectrum of the bilayer bag measured with standard Raman configuration; (c) pure spectrum of Na₃PO₄.

2.3

Large sampling area for more representative, repeatable results of solid samples

Representative sampling of heterogeneous material to obtain bulk properties can be difficult by Raman spectroscopy, which typically has a sampling area of no more than 100 μm in diameter. In multicomponent solid samples, uniform mixing of the ingredients on the scale of 100 μm is difficult to achieve, especially when the grain size is this magnitude or greater. When solids are not well mixed on a scale smaller than the Raman sampling area, collection of spectra at different points on the sample can produce widely varying results due to the sample heterogeneity. This requires many measurements to be made and averaged for representative results for the bulk. A larger sampling area is needed to get more consistent measurements of typical products with multiple components. The difference in results from conventional Raman systems with a spot size of about 100 μm vs. the larger sampling area of several mm with the STRaman probe can be seen in the spectra of a tablet of the over-the-counter pharmaceutical analgesic Excedrin® Migraine containing three active ingredients. Forty three spectra were collected over randomly chosen positions using a confocal fiber probe, which has a sampling spot diameter of roughly 0.3 mm, with fifteen representative spectra overlaid in figure 4 (a). The Raman spectra of the three active ingredients — acetaminophen, aspirin and caffeine — can be detected in all the spectra, though their relative intensities vary greatly, reflecting the uneven distribution of the actives on this size scale. This heterogeneity makes it hard to reliably identify the product. This can be demonstrated by taking a number of such spectra, computing their average and using it to represent the product in a spectral library. As shown in figure 4(b), using the conventional Raman with the 100 μm sampling sport size there is a large variation in the hit quality index (HQI) against the reference spectrum in the library, reflecting the sample heterogeneity at this small scale. (The HQI is a measure of similarity based on the cross-correlation of a measured spectrum to a library spectrum, with a higher value indicating greater similarity). Using a passing threshold of 95, 18 of the sample spectra result in false negatives.

Figure 4:

Comparative spectra of Excedrin migraine tablet as measured at various sample positions. (a) Overlay of 15 spectra acquired with conventional Raman configuration; (b) histogram of 43 HQI values with conventional Raman configuration. (c) Overlay of 15 spectra acquired with STRaman configuration; (d) histogram of 17 HQI values with STRaman configuration.

The average and standard deviation of the 43 HQI values are 93.4 and 6.1, respectively. Using the STRaman approach with a spot size of about 4 mm, each measurement is more representative of the bulk tablet properties. With the average and standard deviation of the HQI values at 99.89 and 0.07, respectively, the reproducibility of the spectrum collected with the larger area, and therefore the identification is greatly improved, with no false negative results in the 17 measurements (Figures 4(c) & (d)).

2.4

Lower power density enables measurements of dark and sensitive samples

Dark material absorbs light and can be a challenge to analyze with Raman without damaging the sample. With the design of the STRaman, laser power is distributed over a larger sampling area, as the laser power is defocused over a larger spot and thereby avoiding sample damage. With this system design, Raman technology is applicable to screening of sensitive samples such as photolabile or thermolabile material, biological tissues, historical art works or archaeological materials. Energetic materials such as gun powder must be very carefully handled and measured in very small quantities with conventional Raman because of the potential of ignition with the focused energy of the laser excitation. The defocused, large area design allows for collection of black gun powder spectrum at high laser power, as demonstrated in figure 5. Raman peaks of the gun powder component sulfur at 217.6 cm^-1 and 471.5 cm^-1, and nitrate at 1049.3 cm^-1 as marked are clearly distinguished.

Figure 5:

Raman spectrum of gun powder measured with 785-nm STRaman technology.

2.5

Versatility of Sampling

An additional advantage of this system design is the versatility of the system. The reflective cavity sampling adaptor can be configured for surface measurement when microanalysis on a sample surface is required. In this way the system is not limited to the large sample area, defocused power and beyond-surface measurements. It can be adapted to a confocal configuration. It can also be configured for confocal Raman with other adaptors allowing standoff, contact or microscopy analysis.