Desenvolvimento de procedimentos para análise inorgânica de biodiesel usando técnicas espectroanalíticas

Abstract

In this work, analytical procedures are proposed for the determination of major and trace elements in biodiesel samples employing flame atomic absorption spectrometry (F AAS), inductively coupled plasma optical emission spectrometry (ICP OES) and inductively coupled plasma mass spectrometry (ICP-MS). Biodiesel samples used in this work were produced from different oleaginous sources (African palm, castor beans, palm, soybeans and an unknown oleaginous). Samples were prepared as microemulsion composed of biodiesel, Triton X-100, HNO3 20 % (v v-1) and n-propanol in the following proportion, 10:5:5:80 (v v-1), respectively. Calcium, Mg and Zn determinations were carried out by F AAS and two nebulization strategies were evaluated, conventional continuous nebulization (CN) and discrete aspiration (DA). Relative standard deviations (RSD) using DA were lower than 4.0, 2.2 and 5.7 % (n = 10) for Ca, Mg, and Zn, respectively. Matrix effects were evaluated by standard additions methods and external calibration using mineral oil to simulate the matrix of biodiesel and inorganic standards. The accuracy of the procedure was checked using addition and recovery experiments for different biodiesel samples and recoveries ranged from 90.8 to 115 %. Moreover, sensitivities reached by using DA for Ca, Mg and Zn were 2.5, 6.1, and 7.0-fold higher, respectively, than those obtained when employing CN. The introduction of microemulsions in ICP OES for Ca, Mg, Na, K, P, and S determinations and the use of oxygen in the composition of auxiliary gas of the plasma to avoid and/or correct matrix effects caused by the high carbon load due to the introduction of biodiesel microemulsions were evaluated. The oxygen gas and nebulization gas flow rates were optimized by measuring intensity signals and signalto- background ratio (SBR) and were fixed at 0.16 and 0.70 L min-1, respectively. Figures of merit were investigated and a good linear correlation was observed (r > 0.99) for different emission lines monitored. The accuracy was evaluated by addition and recovery experiments using different biodiesel samples by monitoring different emission lines and recoveries ranged from 103.3 to 117.5, 93.5 to 113.5, 76.3 to 92.5 and 93.3 to 116.1 % for Ca, Mg, P and S, respectively. The determinations of Na and K were not accurate probably due to the high carbon load content of microemulsions and the high background-signals in the vicinities of their emission lines. xiii Finally, the determination of trace elements in biodiesel samples (Cd, Co, Cu, Mn, Ni, Pb, Ti, and Zn) employing ICP-MS and microemulsions was evaluated. Analogously to the developed procedure using ICP OES, the addition of oxygen gas to the composition of auxiliary gas of the plasma was evaluated and proper performance was attained at 37.5 mL min-1, 4.4-fold lower than the flow rate optimized for ICP OES measurements. Limits of detection employing oxygen gas in the composition of the plasma were higher than without using it, as well as the sensitivities (slopes of analytical calibration curves) were lower when oxygen gas was added to the plasma, probably due to the highest formation of oxides. However, despite oxides formation best accuracies were reached when employing oxygen gas.