Toward a selective detection of ethanol by perspiration

An average proportion of 1% of total alcohol consumed by humans is eliminated through the skin, thus causing the increase of ethanol vapor concentration emitted by the skin [1]. However, one of the major interferences of ethanol detection on the skin is the acetone. Skin acetone is generated from a natural metabolic intermediate of endogenous lipolysis in human and is considered as biomarker of ketotic state of diabetic [2]. Here, we propose to improve the ethanol selectivity of our tin dioxide sensors by using multivariate analysis techniques such as the Principal Component Analysis (PCA). This paper describes the rapid and accurate identification of different compounds such as ethanol, acetone and humidity due to this method in order to recognize ethanol in perspiration.