2, CEA, LIST, Diamond Sensors Laboratory, Gif-Sur-Yvette, , France
3, LeRoy-Eyring Center for Solid State Science, Arizona State University, Tempe, Arizona, United States
Raman spectroscopy is known as a method of choice for the analysis of carbon materials and carbon nanostructures, allowing, among other, the identification of the type of bonding and estimations of the size of coherent domains. This method is now widely used for the characterization of detonation nanodiamond (DND) of various sources.
The Raman spectrum of purified DND usually consists of several characteristic features: (i) the first-order Raman mode of the cubic diamond lattice which is broadened and red-shifted by about 3-8 cm-1 compared to bulk diamond, (ii) broad features with two apparent maxima in the 500-1250 cm-1 range, and (iii) a broad asymmetric line peaking in the 1600-1650 cm-1 (hereafter named “1650 cm-1 peak”) range, depending on the sample origin and its purification. Depending on the excitation wavelength, another feature peaking at about 1750 cm-1 is more or less clearly identified. This Raman spectrum does not strongly depend on the origin of the nanodiamond. In this study, we will focus on the 1650 cm-1 broad band. Indeed, there is still no true consensus for its origin.
For such a purpose, annealing under different atmospheres may be used to modify the surface properties of DND. These treatments confer different surface properties, in particular in terms of surface terminations and surface charge. These treatments are also usual starting points for further complex surface modifications. The modifications induced by different gases can be divided into three categories: in reductive atmospheres, e.g. hydrogen; in an oxidative atmosphere (air, oxygen) and in inert atmospheres: argon or in vacuum. Isotopic labelling (deuterium, 18O) was also considered. In specific cases, high resolution TEM images allowed a better understanding of the observed line shapes.
New experimental conditions were adapted to minimize laser-induced effects on DND and to increase the signal to noise ratio of the spectra, in particular when using excitation wavelengths in the deep UV range (325 nm). However, even for the lowest power levels, subtle signal modifications versus time are still observed, especially for the longest exposure durations. Thus, in this study, we also address the problem of the stability of DND under the UV excitations used for the Raman analysis of detonation nanodiamonds.