Visual binary testing of methanol contained in ethyl alcohol

Abstract

A method for control of methanol traces in rectified ethyl alcohol and alcoholic drinks based on visual binary testing using one reference sample was proposed. An indicator reaction of formaldehyde interaction (product of methanol oxidation) with chromotropic acid disodium salt was chosen for methanol screening. The conditions of indicator reaction proceeding are analogous as for the spectrophotometric technique of methanol determination: methanol was oxidized to formaldehyde with potassium permanganate in an acidic medium; the formaldehyde then reacts with chromotropic acid in the presence of hot concentrated sulfuric acid and forms a violet product (color of this product is stable for 12 hours). It was established that the absorption spectrum of the reaction product does not change on going from 96% ethanol to aqueous-ethanol solutions with a volume fraction of 40% ethanol. The maximum light absorption of the reaction product corresponded to 570 nm. All further studies were carried out in water-ethanol solutions with a volume fraction of ethanol of 40%.

According to regulatory documents the normalized limiting content of methanol (c_lim) in ethyl alcohol of the “Lux” grade (the most common in the alcoholic industry) and alcoholic beverages is 0.01% by by volume counted upon anhydrous alcohol. The comparison sample (the solution of colored reaction product of indicator reaction) had to be less than the normilized level on the value which providing the risk of false-negative test result not more than 5%.

To determination the threshold concentration of methanol in the comparison sample was applied the statistics of observation. For the aim the solution of colored product corresponding to the normalized limiting methanol concentration c_lim = 0.01% by volume was prepared and comparison samples with lower methanol concentrations were also prepared. The interval of unreliability was discovered with the help of observers. The frequency of detecting of the difference in the color of comparison samples and normalized sample (P(c)) changed from 0 to 1 in this interval. The value of methanol concentration 0.0072% by volume counted upon anhydrous alcohol was taken for the lower boundary of the interval and the value of methanol concentration 0.01% by volume counted upon anhydrous alcohol was chosen the upper border of the interval.

This interval was divided on eight concentrations with step Dс = 0.0004% by volume. Three  parallel series of solutions were prepared and 48 observations for each concentration were received. The experimental efficiency curve obtained was checked for compliance with the mathematical functions of the known distributions: normal, logistic, lognormal, exponential and Weibull distribution function using the statistical criterions c² and Kolmogorov-Smirnov λ. The efficiency curve was described by the theoretical functions of the lognormal and Weibull distributions. Calculated at a confidence level of 0.95 estimation of the threshold concentration for the comparison sample was 0.0073% by volume fraction corresponding to anhydrous alcohol. The visual binary testing of methanol trace in alcoholic drinks was carried out. The accuracy of visual binary testing of methanol was confirmed by gas chromatography.