Expected outcome and interpretation of results. The coating of MNM may consist of one or more organic compounds and/or homologous series of compounds. The observation of small very volatile compounds may be due to degradation of larger and/or polymeric compounds. In the following the procedure is illustrated by a case example on nano-sized silver particles (nano-Ag) which were shown to consist of 85 % (w/w) silver by XRF and to be associated 15 % (w/w) organic compounds by TGA (Xxxxxx et al., 2013).These results were in accordance with the information on surface coating of the nano-Ag given by the manufacturer. Without knowing better the first extraction method applied was PLE with methanol at 200 °C. The extract was centrifuged to remove all residual nano-Ag which was not filtered off in the PLE process. The supernatant in different volumes (1-40 μl) was injected onto adsorbent tubes containing Tenax TA and analyzed with TD-GC-MS. This showed that the extract only contained one compound, namely 2-pyrrolidone. The idea that 2-pyrrolidone might be a thermal degradation product emerged and consulting the literature confirmed that both 2-pyrrolidone, and N-vinyl-2-pyrrolidone may be thermal degradation products of polyvinylpyrrolidone (PVP) and that PVP often is used in synthesis mixtures of nano-Ag. Now we had to prove that the coating consisted of PVP. For this purpose the nano-Ag was subjected to ultrasonic extraction in methanol at ambient temperature. The extract was centrifuged to sediment all nano-Ag and the supernatant analyzed by MALDI-TOF-MS (see Figure 1).
Expected outcome and interpretation of results. In the following the procedure is illustrated by a case example on f-CNT functionalized only with carboxylic acid (COOH) groups and thus only releases CO2. When the investigated f-CNT contain enough COOH groups to produce CO2 sufficiently to get above the detection limit a peak may appear in the mass trace m/z 44 as shown in Figure 8. Figure 8 Example of TGA-MS analysis of F-CNT intentionally functionalized with COOH. The upper panel shows MS traces (TIC and CO2 (m/z 44)) and the lower panel shows the TGA curve. The way to derive the amount of CO2 produced by the CNT sample is to transfer the time range for CO2 release detected by the MS (approximately 3 – 28 min in the upper panel in Figure 1) to the TGA curve (marked with green vertical arrows in lower panel in Figure 1) and calculate the weight loss during this period. In the present case this amounts to 0.37 mg CO2. Assuming that no other compounds are released in this period all weight loss can be attributed to release of CO2. Furthermore assuming that all released CO2 originates from COOH groups the content of COOH in the f-CNT amounts approximately 0.28 mmol/g f-CNT. The prerequisites for this result is that no other compounds than CO2 is released in the actual time range and that all O released as CO2 is bound in COOH groups. None of these prerequisites holds in reality but the question is to what extent it is true. The conclusion is that using the described method the result may be reported as maximal content of COOH groups in the f-CNT sample is 0.28 mmol/g f-CNT or < 0.28 mmol/g f-CNT. To get the exact mass of CO2 that is released the MS has to be calibrated for CO2 with a pure solid that release CO2 at thermal decomposition e.g. CaCO3. The method may be supplemented by other analytical techniques that can measure carboxylic acid groups. The simplest method is acid-base titration but this method is not very precise (about ± 50%) (Xxxxxxx et al., 2015). A more advanced method is XPS which can distinguish and quantify all types of bound O.
