International trade in spices has continued to thrive over thousands of years. Present-day producers and importers need to be aware of any legal requirements relating to food safety and quality standards. In addition to pathogens and impurities, the level of metal contaminants present in spices is a further product safety issue. Spices can be contaminated with metals during the growth cycle of the plant or during processing and packaging. Food fraud is another problem, as competitive gains can be made through intentional counterfeiting, substitution, adulteration, or mislabeling/misrepresentation of ingredients within products. Many of these tricks go unnoticed by consumers and regulating government agencies due in part to the lack of standardized methods for the identification of geographic origin.
High-value foods such as wine, rice, oils, honey, fruit juices, tea, coffee, and spices that are marketed according to their provenance are susceptible to food fraud. Additional profits can be made in several ways, for example by blending good quality, authentic products with inferior and cheaper ingredients or by deliberately misbranding a low-quality product as one with a higher value.
Elemental fingerprinting can help combat fraudulent activity. Foods can be authenticated based on the pattern of their trace element content, which is characteristic of the soil composition in the region of production—one study used multi-elemental profiling of 29 tea samples through inductively coupled plasma-mass spectrometry (ICP-MS) for authentication purposes. A similar elemental profiling approach was used in this study to identify the origin of 55 spices from various countries and to differentiate between assorted spices produced in the same country.
Versatile Multi-Element Analysis
Many food testing laboratories already use ICP-MS for the quality control of their products. It is a well-established, fast, multi-element technique to determine a wide range of elements present in a sample at different concentrations. However, given the variety of food types, many foods contain a complex or variable matrix that can give rise to the formation of polyatomic interferences in the ICP-MS spectrum that can affect the accuracy of the data for some elements. A series of recent developments has enhanced the matrix tolerance of ICP-MS and control of polyatomic interferences, improving its suitability for the analysis of foods. ICP-MS equipped with an ultra-high matrix introduction system enables the plasma to tolerate samples containing up to 25 percent total dissolved solids (TDS). Octopole-based collision/reaction cell (CRC) technology removes polyatomic interferences arising from the plasma and sample matrix using kinetic energy discrimination (KED) with a single gas (helium mode), improving the data quality of foods with complex matrices.
Over 50 spices from around the world of known origin were received from a business-to-business spice supply company based in the U.S. Knowing the origin of samples is critical to the development of a reliable model that can be used to authenticate unknown samples. All spice samples were microwave digested in acid (MARS 6, CEM).
Since our lab is equipped with both ICP-optical emission spectroscopy (ICP-OES) and ICP-MS, we first used ICP-OES as a screening technique to establish the concentration levels of elements present in the spice sample digests. The same samples were then analyzed using ICP-MS. Rather than dilute the samples to bring the high-level elements (aluminum, calcium, germanium, potassium, magnesium, sodium, phosphorus, sulfur) into range, we used the ICP-OES results for these elements in the statistical analysis.
A 7900 ICP-MS and 5110 ICP-OES (Agilent Technologies) fitted with an SPS 4 autosampler (Agilent Technologies) were used to analyze various spice samples. Mass Profiler Professional (Agilent Technologies) chemometric software was used for statistical analysis of the data set.
Validating the Analytical Method
To verify the spice sample digestion process, three National Institute of Standards and Technology (NIST) standard reference materials (SRMs) were analyzed by ICP-MS and ICP-OES. The mean concentrations (ppm) of three repeat measurements of three SRM digests were in good agreement (80-120 percent) with the certified concentrations, where certified concentrations were provided.
A spike recovery test was then carried out to check the accuracy of the elemental method for spice sample analysis. Four random spice samples were spiked with all elements at 20 and 60 ppb and measured using ICP-MS and ICP-OES. The quantitative results for the spice samples showed that the concentrations of aluminum, potassium, calcium, magnesium, sodium, iron, phosphorus, sulfur, silicon, zinc, and manganese were relatively high in all four spice samples. The spike results for these elements were therefore invalid as the spike levels were too low (20 times lower) relative to the levels present in the unspiked samples. The recoveries for all remaining elements were within ±20 percent.
All spices were analyzed, and the multi-element data batch file (55 spice samples, nine replicates) was imported into MPP chemometric software for statistical analysis. Principal component analysis (PCA), an unsupervised technique, was used to find the direction of the greatest variance in the elemental data and display the samples based on these differences and similarities. As shown in Figure 1, the spice samples were separated fairly well based on country of origin and by spice.
Overall differences between the elemental composition of spices from the 13 different countries were found. As seen in the PCA (see Figure 1), spice elemental profiles were found to discriminate country of origin and explain 47.22 percent and 11.69 percent of the variance in PCA components 1 and 2, respectively; however, the countries were not completely separated.
We were interested to see if origin could be distinguished when examining one spice. This can be demonstrated in the PCA of rosemary, where clear separation between samples from Morocco and Tunisia is shown (see Figure 2). In addition to discriminating between countries, we saw in Figure 1 that elemental profiles could also distinguish some spices. We further investigated whether spices originating from one country could be separated. The PCA in Figure 3 shows the elemental composition of multiple spices within Egypt and Turkey. Clear separation was seen between four spices from Turkey, and possible spice discrimination was seen in samples from Egypt.
Initial Findings and Future Aspirations
ICP-MS can be used for the quantitative analysis of the widest range of elements in spice samples, producing large data sets for statistical analysis. ICP-OES can also be applied for elemental fingerprinting studies using data for all but the lowest concentration elements.
Exploratory data analysis using PCA showed that the elemental composition of spices is influenced by the country of origin, allowing discrimination between 13 countries. Four different spices from the same country were also separated using the methodology, as was the same spice from two different countries.
More samples are needed to strengthen and test the fingerprinting model to authenticate spices. However, because of current tracking issues, obtaining spices of known origin is challenging. Once established, the method could form a valuable part of a food manufacturing and distribution facilities’ food fraud program—potentially with economic- and health-related benefits for the consumer.
Dr. Nelson, an assistant adjunct professor for viticulture and enology at University of California Davis, is a market development spectroscopy scientist at Agilent Technologies. Reach her at firstname.lastname@example.org. Tanabe is a graduate student of agricultural and environmental chemistry in the Department of Viticulture and Enology at UCD. Gilleland and Whitecotton are application engineers at Agilent Technologies. And Hasty and Anderson work for CEM Corp.
*For enlarged Table, CLICK HERE.