Detecting Adulterated Meat with LIBS

Economically motivated food adulteration is a problem around the world. The 2013 horse meat scandal in Europe, in which horse meat was substituted for beef and pork, did not cause serious health problems, but certainly upset many consumers and led to a desire for increased quality control over meat products. Ismail Hakki Boyaci of the Food Research Center at Hacettepe University, in Ankara, Turkey, and his team, including Gonca Bilge and Banu Sezer at Hacettepe University, Murat Velioglu at Namik Kemal University, and Efe Eseller at Atilim University, have developed a method using laser-induced breakdown spectroscopy (LIBS) to detect meat adulteration. Boyaci recently spoke to us about this work.

You have developed a method for identifying meat species using LIBS, to detect food fraud. Are existing methods insufficient?

Meat products have an important role in human diet because of their high nutritional content. However, the high price of meat makes it an attractive product for adulteration to make a profit. Meat adulteration not only leads to financial, ethical, and health problems, but also raises concerns about religious beliefs and may cause allergic reactions. Therefore, meat adulteration is a common problem in many countries, and detection of meat adulteration has become crucial. Hence, the meat industry urgently needs a rapid, practical, and reliable method. We developed a new method that is based on elemental composition differences between meat species to identify meat species using LIBS.

At present, there are many analytical techniques for identifying meat species, including polymerase chain reaction (PCR), real time PCR, gas chromatography-mass spectrometry (GC–MS), high performance liquid chromatography (HPLC), isoelectric focusing, capillary gel electrophoresis, and enzyme-linked immunosorbent assay (ELISA) techniques. Although genetic methods are quite sensitive and reliable, they are expensive techniques that require specialists and DNA and protein extraction steps. Chemical methods are also time-consuming. Due to cross reactions caused by antibody biomarkers, immunological methods may give false results. Owing to the disadvantages of these techniques, the meat industry urgently needs a new, rapid, in situ, accurate, and sensitive method that will identify the animal species and quantify the adulteration ratio.

The sample preparation process for this method comprises many steps, including drying, defatting, grinding, and forming the powdered sample into a pellet. Is this a typical process for analyzing biological tissue samples via LIBS? Did you have to adjust the process to optimize the analysis?

In theory, LIBS provides rapid analysis in solid, liquid, and gas matrices. However, it is known that water and fat content decrease the intensities of atomic emissions. To overcome these difficulties, dried samples have been used in many studies. We also used dried, defatted, grinded, and pellet-formed samples in our study to achieve the best results. This is an important sample preparation procedure. However, system optimization and elimination of the sample preparation step or shortening sample preparation time is possible by using a high-energy laser source or the double-pulse LIBS technique, which is successful in even liquid analysis. Our laser energy was not sufficient for ablation of fat and water in meat samples; thus we analyzed the meat samples and optimized the methods based on the properties of our LIBS system. This study only presents the potential applicability of LIBS for determination of meat adulteration. In the next step of the study, we will try to eliminate sample preparation steps.

What differentiating factor among meat species can LIBS identify?

Mineral differences between meat species are differentiating factors for identifying them. There are many studies in the literature that show that Ca, Mg, K, Na, Zn, Cu, and Fe compositions vary between meat species. However, the differences between them have not been used for classification in the studies carried out so far. The elemental composition of meats can vary with species, season, region, and parts of the animal. Therefore, in this study, several animals from different regions and different parts of these animals for each species were used to compensate for those variations. In conventional methods, DNA and protein analysis are used for identification of meat species. Using mineral differences for classification is a new approach.

What is the role of chemometrics in making LIBS an effective technique for identifying meat species?

Our LIBS data are very complex because of the matrix of meat samples, which contain many elemental bands between the 350 and 850 nm wavelengths. In this study, identification of meat species was performed by evaluating the whole LIBS spectrum of the elemental composition of meat samples. Therefore, multivariate data analysis techniques were required. Chemometric techniques such as partial least squares (PLS) and principal component analysis (PCA) are widely used to enhance the analytical performance of LIBS. These advanced techniques reduce the complexity of spectra and provide robust and accurate models by reducing the dimensions of variables. They also eliminate laser fluctuations from shot to shot as well as physical and chemical matrix effects. In this study, PCA was used to discriminate the three different meat species (pork, chicken, and beef) and PLS was used to determine the adulteration ratio.

What results have you obtained so far? Is LIBS effective for this application? How does the performance of LIBS for meat identification compare to currently used techniques?

We succeeded in making qualitative and quantitative determinations of pork, chicken, and beef samples. Qualitative discrimination of meat species was performed using LIBS combined with PCA, as can be seen in Figure 1. Good discrimination was achieved between beef, pork, and chicken. In particular, the discrimination between beef and pork is very clear. Although the samples belong to different parts of animals from different regions, the method correctly identified the species.

Figure 1: PCA graph of meat species analyzed using LIBS.

Furthermore, we were able to determine the adulteration ratio. For this purpose, we prepared adulterated chicken and pork beef samples at 10–50% ratios to obtain a calibration graph and analyzed the samples with LIBS combined with PLS. We achieved the determination of meat adulteration above 5% with a 12% relative standard deviation (RSD). Meat adulteration is performed to make profit, and therefore the adulteration ratio is generally expected to be more than 10%. That means that a 5% detection limit is desirable.

It can be said that LIBS is an effective method for meat identification, but obviously it needs further development. This study has only shown the potential applicability of LIBS in this field.

According to the literature, PCR is considered to be the most reliable method for determining meat adulteration. However, the results of the PCR method are expressed as genome/genome, and there is no evidence of a correct correlation from DNA results to meat content (w/w) in unknown meat products. There are some correlation problems in DNA-based methods, such as differences in species genome size, tissue cell size, DNA extractability, DNA degradation, mitochondrial distribution, and fat and water content (cell number per gram of meat). This is a problematic issue in the quantification of meat species and calculating the limit of detection (LOD). Also, conventional methods require a DNA and protein extraction step, a long analysis time, and specialized personnel.

The LIBS methods, however, can provide the results as w/w. In addition, although the mineral content of meat can vary depending on the part of the animal and its geographical origin, these differences are negligible compared to the differences between meat species, and thus do not cause significant changes in the data obtained using LIBS.

What are your next steps in this work?

Because of the limitations of the LIBS instrument we used in this study, we had to apply some sample preparation steps. For our future studies, we aim to eliminate sample preparation steps by using a LIBS instrument with a high-energy laser source or double-pulse capability, since those features make it possible to analyze samples having a high water and fat content. Another point is the variety of meat species and the number of specimens. We used only three meat species in this study-chicken, pork, and beef. In future studies, we will use LIBS to analyze other meat species such as horse, lamb, and turkey, and will analyze a higher number of samples. Also, in our studies we used minced meat, but meat adulteration is generally performed in meat products. Therefore, the applicability of this method in meat products should also be investigated.