RAPID•ID™ Eliminates the Costly Misidentification of Similar Plastics
Share This Page
•
• Free Access to 80,000 Datasheets |
|||||
|
RAPID•ID uses Raman spectroscopy to accurately identify minute differences in plastics in a matter of seconds. RAPID•ID is a molecular analysis tool that quickly and easily scans the material for all the covalent bonds on a sample, and detects differences in the chemical make-up, even if very small. This lightweight, palm-sized tool then compares the results from any tested sample against a set of library spectra in memory. Identifications are determined using a differential correlation matching algorithm. When the materials are too similar, the user can choose three different methods to enhance the matching algorithm. First, the algorithm can be taught to examine particular regions of the spectra and determine the presence or absence of peaks; second, users teach the algorithm to calculate the ratio of peaks in pre-determined locations and set thresholds for positive or negatives matches, and, third, the user can set the algorithm to emphasize the correlation in a specific sub-region or regions (e.g. “blocks”) of the spectra to confirm or negate the closeness of the match to a particular reference. These methods can be used separately or together in tandem. Following is a description of all three methods. 1. Presence / Absence of Peaks Chemically, both products are virtually identical and can be very difficult to differentiate using standard methods such as burning and sniffing. However, RAPID•ID with Raman technology can detect the difference even when their chemical composition appears to be the same. A peak is clearly visible in the 1,621cm-1 region (Figure 1). A standard correlation algorithm will yield very high correlations between cast and regular acrylic, of 96% or higher, between the two types. This is because most of the region of interest in both spectra is basically identical. With the use of the secondary matching algorithm, RAPID•ID can easily separate the two materials to achieve a 99% or higher accuracy in tests using the “Find peak” algorithm. 2. Peak Ratios Since the differences are driven by the crystalline structure, branching and cross-linking of the polymer chains, RAPID•ID with Raman spectroscopy will detect those differences, as the total count of covalent bonds will vary accordingly. Specifically, the shape of the peak centered around 1,441cm-1 is an excellent indicator of the degree of crystallinity of the material. The spectra for LDPE, HDPE and UHMWPE are very highly correlated to each other, at 91% or higher (see Figure 2 and Table 1).
This may increase the possibility of false positives or negatives when testing one against the other. By calculating the ratio of the main peak at The peak ratios can be used to differentiate between the three polyethylene samples in this study (see Table 2). There is no overlap of the ranges, which results in very high accuracy. The ratio of the peak centered around 1296cm-1 to the one centered around 1420cm-1 can also be used as a benchmark.
3. Block Matching
To illustrate the recipe differences, samples of four different PVC suppliers were tested. The samples from suppliers A & B are visibly similar to each other. However, supplier B’s product was reverse-engineered from supplier A. Both are clearly different from the samples from C and D, which also are very similar to each other (Figure 4). The correlations between these four samples indicate differences (Table 3). An extrusion company that was using supplier C material was experiencing difficulties with the processing characteristics of supplier D’s material, and they needed to understand if there was a difference between the two. Using the standard matching algorithm, RAPID•ID can distinguish between the two groups (A and B versus C and D). However, a more advanced matching method is required to differentiate between the two suppliers in each group. In this particular case, a closer look at the overlapped spectra of the samples from supplier A and B as well as C and D shows that there is a significant difference in the wavenumber region from 980 to 1,050cm-1 (Figures 5 and 6) as well as the region from 800 to 925cm-1. An advanced library was set up that would require a minimum matching threshold of 90% correlation for each block before returning a positive result. The block matching differentiates between samples C and D, as the region from 800 to 925 cm-1 only showed a correlation of 69%. (Table 4). However, this region was not as good a benchmark as expected, as both A and B were highly correlated in that block, at 92%.
This required a third condition to be added to the library, namely a test for the ratio of the peaks at 1,000cm-1 and the one at 1,433cm-1. After addition of this rule, the accuracy of the overall test was increased to 99%. Count on RAPID•ID to meet your most complex plastics ID challenges For more information about how RAPID•ID can quickly and accurately identify even the most minor differences in the make-up of plastic materials while lowering your costs, call 1-866-301-6328 or email rapidid@deltanu.com
Join the IDES Network: Email Alerts • Blog • Twitter • LinkedIn |
|
||||
polymerized and pelletized and then used in most common molding processes. When high-quality PMMA sheets are required, they can be produced by cell casting, in which case the polymerization and molding steps occur concurrently. These sheets tend to have much higher strength and the price difference between cast and regular acrylic can be as much as five times. 
1,441cm-1 to the one at 1,420cm-1 (Figure 3), and implementing it as a secondary test method, the accuracy of the test can be reliably increased to 99% or higher. 
