The extraction of trimethylamine (TMA) with toluene was found to be affected by temperature, sorts of alkali and number of shaking times, as shown in Fig. 1 and Table 1, whereas the color development of TMA with picric acid was not influenced by temperature. From these observations, the extraction procedure was modified to be carried out at a definite temperature (30°C). In an incubator, test solutions, reagents and funnels were previously warmed before extraction and the funnels shaken 60 times were again left for 5 minutes in it before the separation of toluene layer.
In the second place, the same volume of 25% potassium hydroxide solution was adopted instead of the original 50% potassium carbonate solution, which was confirmed to be not suitable for the TMA solution resulting from the reduction of trimethylamine oxide (TMO) with Devarda's alloy and hydrochloric acid. Figures in Table 2 indicate that some constituents of the alloy may disturb the extraction of TMA in the case of potassium carbonate, but not in the case of the hydroxide. This agrees well with the fact that the carbonate forms white colloidal precipitates, perhaps carbonates of Al and Zn, in the aliqout of the heated reducing mixture, and the precipitates were suspected to adsorb or occlude TMA. The hydroxide does not give such precipitates.
The modification above-mentioned facilitated the determination of the smaller amount of TMO and the modified method gave over 97% recovery down to 0.025mg N which is far lower than 0.4mg N in the DYER method (Fig. 2). The recovery of TMO added to carp muscle extracts was also satisfactory (Table 3).