Hatanaka, Rafael RodriguesSequinel, RodrigoGualtieri, Carlos EduardoTercini, Antônio Carlos BergamaschiFlumignan, Danilo Luizde Oliveira, José Eduardo2014-05-272014-05-272013-03-14Talanta.0039-9140http://hdl.handle.net/11449/74827Lubricating oils are crucial in the operation of automotive engines because they both reduce friction between moving parts and protect against corrosion. However, the performance of lubricant oil may be affected by contaminants, such as gasoline, diesel, ethanol, water and ethylene glycol. Although there are many standard methods and studies related to the quantification of contaminants in lubricant oil, such as gasoline and diesel oil, to the best of our knowledge, no methods have been reported for the quantification of ethanol in used Otto cycle engine lubrication oils. Therefore, this work aimed at the development and validation of a routine method based on partial least-squares multivariate analysis combined with attenuated total reflectance in the mid-infrared region to quantify ethanol content in used lubrication oil. The method was validated based on its figures of merit (using the net analyte signal) as follows: limit of detection (0.049%), limit of quantification (0.16%), accuracy (root mean square error of prediction=0.089% w/w), repeatability (0.05% w/w), fit (R 2 =0.9997), mean selectivity (0.047), sensitivity (0.011), inverse analytical sensitivity (0.016% w/w-1) and signal-to-noise ratio (max: 812.4 and min: 200.9). The results show that the proposed method can be routinely implemented for the quality control of lubricant oils. © 2013 Elsevier B.V. All rights reserved.engEthanol contentInfrared spectroscopyLubrication oilValidationArtigo10.1016/j.talanta.2013.02.047WOS:000319089100027Acesso restrito2-s2.0-84874769169