ИСТИНА

Graphite oxide (GO) is an interesting material that can scarcely be referred to as a substance because of its irregular and variable structure and composition. Having been known for more than a century and a half, it still did not unlock all secrets of the nature, mutual arrangement, and even relative amounts of different functional groups that cover graphite layers and determine numerous promising properties of the substance. Even the C/O ratios in samples obtained according to different oxidation techniques differ, not saying about the inevitable nonzero content of hydrating water that may be as high as 10-17%. It is worth noting that arrangement of functional groups on a GO surface partly covered with water molecules, which naturally form hydrogen bonds with the groups, can be considered as a very interesting supramolecular system where all important oxygen-containing building blocks are strongly grafted to a carbon skeleton and interconnected either directly or via H2O molecules. According to XPS and solid-state NMR studies, graphite oxide contains chiefly oxidized carbon in the form of epoxide and hydroxide groups and non-oxidized carbon, as well as a much smaller amount of carboxyls, ketones, and geminal diols. Absorption ir spectra seem to be most informative due to the presence of several clear signals in the fingerprint range and a broad band from 3700 to 2200 cm-1, which can be deconvolved into a series of Gaussians that may give rise to ideas about the probable nature of the functional groups. However, for the same reason as in the case of XPS and NMR, namely the absence of adequate tabulated reference data, the spectra cannot be interpreted if no additional information is used. The source of such information is quantum chemical modeling. We have carried out nonempirical simulations of model С54Hp(O)k(OH)l(COOH)m(H2O)q systems that represent hypothetic fragments of an oxidized carbon sheet at an assumption about different kinds of prevailing functional groups and their combinations. Calculated absorption spectra of the systems (based on the results of the normal-coordinate analysis) were visualized with the use of Lorenzian broadening and compared to the recorded infrared absorption spectrum of the synthesized specimen. The comparison enabled us to formulate certain conclusions about the nature of spectral signals in the fingerprint range. In particular, it became clear that there are mutual effects of different functional groups, solvating water molecules, and carbon skeleton, which can by no means be predicted without theoretical consideration.