Raman spectrum-cluster analysis method was developed for rapid discerning of adulterated olive oil using multiple batches of olive oil, soybean oil, corn oil, rapeseed oil, palm oil, cottonseed oil and refined swill-cooked dirty oil with various origins and brands as samples. Raman spectrum patterns of olive oil, low-price edible vegetable oil and refined swill-cooked dirty oil were studied under the conditions of 780 nm and 532 nm laser sources with common gratings and 532 nm laser source with extended grating. Then the adulterated olive oil was discerned by cluster analysis method. The results indicated that the extended and first-order derivative Raman spectra showed abundant information and significantly different spectrum patterns among olive oil, low-price edible vegetable oil and refined swill-cooked dirty oil in 532 nm laser source. The cluster analysis model established on full-range spectrum information could discern olive oil as well as kinds of adulterated olive oils. The discriminant rates were 100% for 30 olive oils, 105 low-price edible vegetable oils, 38 refined swill-cooked dirty oils and 75 olive oils mixed with at least 5% refined swill-cooked dirty oil. Besides, the discriminant rates were above 94% and 88% respectively for 180 olive oils mixed with at least 5% low-price edible vegetable oil and 72 vegetable oils mixed with at least 5% refined swill-cooked dirty oil. The measurement process required no sample preparation or chemical reagent consumption, and the time cost was 5 min for each sample, therefore the discerning of adulterated olive oil was rapid, non-destructive and accurate. |