Detection of adulterated proteins in the dairy industry

SUMMARY

One of our customers in the dairy industry tested Chemometric Brain to detect adulterated samples of milk protein. The adulterated samples could be clearly identified when using Chemometric Brain’s model for Milk Protein Concentrate 85%. With this simple and quick technique, producers can discard adulterated raw materials before including them in the manufacturing process, which may lead to economic and reputational losses.

CONTEXT

Food adulteration can have serious effects on health and safety. Nowadays, consumers are more aware of this and demand high-quality food products. Detecting adulterations in a production line, whether intentional or accidental, can avoid big economic losses and reputational risks.

One of our customers in the dairy industry tested Chemometric Brain to detect adulterated samples of milk protein.

TESTED PRODUCT

Samples of several batches of Milk Protein Concentrate (MPC) 85% adulterated with different concentrations of sweet whey protein.

ISSUE

Detection of adulteration proteins in a set of MPC 85% batches.

APPROACH

The customer wanted to test our software before implementing it in his company. To carry out the analysis, four samples from a validated batch were intentionally adulterated at different concentrations: 1%, 2%, 5% and 10% of sweet whey protein. The analysis of the protein in the samples (wet chemistry) did not show differences in the composition with respect to other validated batches of Milk Protein Concentrate 85%.

However, when the samples underwent a Chemometric Brain’s qualitative analysis, the adulterated samples could be clearly identified when using our model for MPC 85%.

The “fingerprints” of the new batches were analyzed and compared with those previously validated in the company through a Principal Component Analysis (PCA). As shown in Figure 1, the adulterated MPC 85% samples (marked in light green) were different from the ones included in the model, since they were outside the mesh that defines the confidence area (marked in dark green).

Chemometric Brain’s software was able to easily detect the adulterated samples where Sweet Whey Protein had been added.

Figure 1 shows the representation of PC1 versus PC6, where the Chemometric Brain model not only separated the adulterated samples from the validated ones but also placed further away from the confidence area those where the percentage of the Sweet Whey Protein was higher.

CONCLUSION

Food adulteration is a critical point in the production line that must be monitored. But detection is not always simple.

In this example of adulterated proteins in the dairy industry, we show how Chemometric Brain can easily identify adulterated samples by performing a qualitative analysis. With this simple and quick technique, the customer can discard adulterated raw materials before including them in the manufacturing process, which may lead to economic and reputational losses.

‍

‍