Determinação de manganês em etanol combustível por voltametria de redissolução

Abstract

The ANP (National Agency of Petroleum, Natural Gas and Biofuels) inspects only the levels of iron and copper in fuel ethanol, however the determination of manganese is also necessary due to the negative effects of metals that accelerate the formation of gums and sediments inside engines and storage tanks. The metal alloys employed since the manufacturing until to distribution of this fuel are composed of carbon steel and / or stainless steel which contain considerable amounts of manganese in their chemical composition. This work proposed the determination of manganese by square-wave cathodic stripping voltammetry employing a glassy-carbon electrode as detector. The voltammetric characterization was accomplished between -1,0 and +1,5 to verify the electrochemical behavior oh manganese. For its determination, the following conditions were employed: deposition potential of 1.2 V for 180 seconds, conditioning potential of -0.5 V for 30 seconds, 0.04 mol L-1 Britton-Robinson buffer solution at pH 4, ethanol ratio in the electrochemical cell of 2.5% (v/v), amplitude of 100 mV, frequency of 30 Hz, a step of 10 mV and stirring speed of 2250 rpm. Some parameters for the validation of method were evaluated, such as repeatability, linearity, recovery test, limits of detection and quantification. The following results were obtained: repeatability with standard deviation of 5.58 % for 10 readings, analytical frequency of 16 analysis h-1, recovery values between 90.8 % and 110.9 %, linear range between 3 and 175 μg L-1, detection limit of 0.9 μg L-1 and quantification limit of 3 μg L-1. The content of manganese in the analyzed fuel ethanol samples were below the detection limit of the proposed method. Searching for the development of an alternative method for the determination of manganese, the fuel ethanol samples were pretreated by complete evaporation followed by acid digestion of the organic residue with an oxidant mixture (H2O2 + HNO3). However, this method did not provide acceptable recovery values (111 160%) employing the same optimized conditions to the use of glassy carbon for the direct analyses of fuel ethanol, which indicates a possible matrix effect resulting from the sample pretreatment.