Once an infrared spectrum is obtained, it needs to be analyzed, a task that requires skill, experience, and the availability of reference spectra. As much as possible should be learned about the sample and its history, including direct observation. Computers equipped with search software and electronic databases of reference spectra make this job easier.
The infrared spectrum is almost the fingerprint of a particular molecule or coating formulation, characterized by the number, shape and intensity of absorption bands. This pattern can be compared to a collection of known patterns, whether manual or computerized. An extensive collection of reference spectra is available in both printed and electronic forms. Nicolet Analytical Instruments and the Bio-Rad Informatics division offer an extensive collection in computer-ready form. Hummel has prepared several excellent collections of infrared spectroscopy useful for coatings applications, including Polymer and Related Materials. The Federation of Coatings Technical Societies has published an excellent book on infrared spectroscopy for the coatings industry, including instructional texts and an excellent Reference spectrum. Spectra are now available in digital form from Nicolay Analytical Instruments.
Computer searches suggest effects of pure compounds or mixtures in the library. The mixture may not give a satisfactory match through the search software. If a computer search fails to find a good match, it may be worthwhile to simplify interpretation by separating the coating into components and scanning each. Even if a computer search fails to find a good match, the presence of group frequencies in the IR spectrum (explained below) usually means that a good match has similar functionality to the unknown group. It's a good idea to scrutinize the conclusions drawn by such software before accepting their legality.
In addition to being a fingerprint, the infrared spectrum also contains important structural information. Infrared absorption bands can be divided into characteristic bands and fingerprint bands. The latter are specific to a particular molecule. Characteristic bands or group frequencies are caused by specific functional groups such as esters or amides. The frequency of these groups is almost independent of the rest of the molecule in which they occur, and they can be used to determine which functional groups in a molecule are present and which are not. Their slight dependence on the rest of the molecule can be exploited to gain additional information. For example, unsaturated and saturated esters have slightly different carbonyl stretch bands. The compilation of group frequencies can be used to guide the analyst through interpretation. Once the functional groups in the unknown substance have been identified and other options have been ruled out, the analyst can search for available reference spectra of compounds or polymers that have the same function as the unknown substance.