Study On The Detection Of Liquid Milk Adulterated With Nonprotein Nitrogen By Mid-infrared Spectroscopy

Cow milk, as well as other dairy products, are affluent in terms of nutrition and have afforded people natural food that akin to breast milk. However in recent years, for interest reasons, adulterated and artificially-rich in protein milk are continuously proliferated among some merchants’good. Hence, it is of immediate significance to identify a detecting method which is fast, easy-to-conduct and keen to numerous adulterants.This dissertation adopts middle infrared spectroscopy to collect spectrums, thereby. conducting feature analysis among a number of adulterants in liquid cow milk including carbamide, ammonium chloride, ammonium phosphate, ammonium nitrate, sodium nitrate, sodium nitrite and melamine. Given the acquired findings, this dissertation proceeds with qualitative differentiation and quantitative in-depth research, main approaches are stated below:(1) Given the spectrums collected from the above seven adulterants, observed physicochemical properties, together with their optical features, are used to ascertain the characteristic absorption peak while excluding alterant impacts from milk, water and second derivative spectroscopy. Results are revealing the fact the each of the adulterant has its typical absorption peaks. The characteristic peaks of carbamide are 1626 cm-1 and 1597 cm-1; the main typical peak of ammonium chloride is 1452 cm-1; the characteristic peaks of ammonium nitrate as same as the sodium nitrate’s are 1348 cm-1; the main typical peak of ammonium phosphate is 1075 cm-1; the sodium nitrite’s is 1232 cm-1 and the main typical peak of melamine is 1556 cm-1. For instance, ammonium nitrate and sodium nitrate are differed in peak form while identical in peak position. Such finding will serve as a cornerstone to manifest adulterants in milk.(2) Matlab is used to construct program for the sake of qualitatively examining adulterants in milk, where adulterants of differed consistency are analyzed. Results have claimed its "yes or no" function while not accurate enough to discriminate them among the mess. Confronted with such situations found in sodium nitrite, sodium nitrate and ammonium nitrate, it is well-advised to look into the characteristic absorption peak again, namely to synthesize conclusions derived from (1).(3) In terms of quantitative analysis, quantifiably-determined characteristic absorption peak is selected from each of adulterants, which are further evaluated using peak area and peak height. Apart from previous methods, specification curves are derived when modeling unary linear regression and polynomial curves. Upon comparisons it is argued that the polynomial model fits better and peak area method applies as well. When using the polynomial model the relative error can within 2% whether peak area and peak height.