Concept
A nuclear magnetic resonance spectra database is an electronic repository of information concerning Nuclear magnetic resonance (NMR) spectra. Such repositories can be downloaded as self-contained data sets or used online. The form in which the data is stored varies, ranging from line lists that can be graphically displayed to raw free induction decay (FID) data. Data is usually annotated in a way that correlates the spectral data with the related molecular structure. The form in which most NMR is described in literature papers. It is common for databases to display line lists graphically in a manner that is similar to how processed spectra might appear. These line list however lack first and higher order splitting, satellites from low abundance isotopes like carbon or platinum, as well as the information concerning line width and other informative aspects of line shape. The advantage of a line list is that it requires a minimal amount of memory. Once an FID is processed into a spectrum it can be converted into an image that usually takes up less memory than the FID. This method requires more memory than a line list but supplies the user with considerably more information. The processed image has less information that a raw FID but it also take less memory and is easily displayed in browsers and requires no specialty data handling software. The raw free induction decay data obtained when performing the experiment are stored according to the formatting preferences of the instrument manufacturer. This data format contains the most information and requires the most storage space. A variety of commercial and free of software programs allow users to process FID data into useful spectra once FID data is downloaded. Some database search methods are commonly available: Compound name — May include official IUPAC names and common names. Molecular formula — Either an exact formula or a range. Molecular structure — This method requires a molecular editor interface. Registration number — Commonly the CAS Registry Number but most databases also have their own numbering scheme.