Using Benchtop NMR as a QA Solution in Food and Agriculture

The measurement of parameters such as oil, fat, and moisture content is important to help define the nutritional value and product quality for many foodstuffs and agriculture products. In particular, these values are widely used to determine energy content and to calculate the proportion of other components in foods (e.g., carbohydrates). In addition, the oil and fat content may significantly affect the texture, perceived quality, and flavor of products. Thus, accurate measurements of the oil and fat content enable manufacturers to achieve higher standards in nutritional characterization and quality control of foodstuffs. Similar considerations apply to the measurement of seeds and feedstocks used in agriculture.

Use of solvent extraction method, instrumental method, and benchtop NMR to measure oil and fat content in foods

Solvent extraction techniques are commonly used for the determination of fat content. However, they tend to be slow, cumbersome, and require highly skilled personnel. In addition, many of the often hazardous chemicals used are becoming increasingly unacceptable according to international environmental standards. Despite these issues, solvent extraction continues to be used as a reference measurement for quality control.

Instrumental methods are often referred to as secondary techniques since they are usually set up to match the results produced by solvent extraction. To provide a result equivalent to the traditional extraction techniques, secondary techniques require a correlation against the reference technique used. Although they are generally fast, instrumental secondary techniques differ greatly in their need for calibration and maintenance. Calibration maintenance, together with consumables, can add significantly to the cost of ownership. For example, although supercritical fluid extraction (SFE) is reasonably fast, it requires high maintenance, and the cost of compressed CO2, used to extract oil, is also significant. Near infrared (NIR) is commonly used for on-line monitoring, but is difficult to apply on opaque samples since it can only scan the surface. It is also complex to calibrate because measurements are sensitive to product granularity and additives such as seasoning; therefore it is difficult to maintain accurate calibrations on a large variety of product types. For that significant reason, NIR has limited applicability for the quality control of oil (fat) content in many foods.

Figure 1 - Calibration for foodstuffs with fat contents of 2.1–40.2% by mass.

Table 1 - Instrument repeatability test results for fat content measurements

In contrast to the standard wet chemistry methods and various alternative secondary techniques, benchtop nuclear magnetic resonance (NMR) provides a fast, direct, and user-friendly method to determine the fat and oil content in foodstuffs. It also has the benefit that the instrument can be operated by a non-NMR expert. The technique is based on measurement of the NMR response obtained from oil (fat) in the product, and quantification of the oil (fat) content by calibration.

Benchtop NMR for food analysis

The instrument can be calibrated to cover a concentration range from 0 to 100% oil (fat), or a simple calibration can be made using a single sample of oil. NMR is very stable over the long term, therefore minimizing recalibration. Sample measurement time is short (typically 20 sec) and there is minimal sample preparation. It has the advantage of being virtually insensitive to sample matrix granularity and additives such as spices, flavors, colors, and salt. Lastly, since the NMR technique is nondestructive, repeatability measurements can be made with ease.

Calibration and instrument repeatability

A calibration can be generated using either one 100% oil (fat) sample or a set of 3–6 samples of real products with predefined oil (fat) contents spanning the range of concentrations of interest. Figure 1 shows a calibration for foodstuffs with fat contents ranging from 2.1 to 40.2% by mass. Dried samples of macaroni and cheese, muesli, garlic bread, powdered milk, and potted meat were used to generate the calibration in Figure 1. As seen in the figure, NMR gives excellent linear correlation between the NMR response and the concentration of fats in the products.

Instrument repeatability was tested by measuring one sample 10 times. After every test measurement, the sample was transferred from the instrument back to the conditioning block for 20 min to be conditioned at 40 °C and then measured again. Table 1 shows the repeatability test results.

Figure 2 - MQC benchtop NMR system.

The Oxford Instruments MQC-23 with 0.55 Tesla magnet fitted with a 26-mm-diameter (10-mL) sample probe (Oxford Instruments Magnetic Resonance Division, Abingdon, Oxon, U.K.) is well-suited for this application. Shown configured in Figure 2 as the Fat Content in Foodstuffs package, it comprises the MQC-23 instrument module, a unit that is controlled by its on-board computer, together with MultiQuant operation software and EasyCal, a suite of subroutines for calibration, analysis, and final result calculation, which is dedicated to the Fat Content in Foodstuffs application. To help the user begin and to ensure ongoing reliability of results, standard test and “setting-up samples” with nominal fat content values of 10, 25, and 40% are provided for calibration maintenance and quality control. Many users also add a dry block heater to keep their samples at a constant temperature (40 °C).

The MQC offers multiple advantages. High signal sensitivity and a small benchtop footprint help to guarantee good performance without the increased cost of ownership of larger-sized instruments, a point also supported by its low maintenance costs.

Measurement of oil and moisture in seeds

Accurate and fast determination of oil content is important to breeders, growers, and buyers for determining the commercial value of oil-bearing crops such as rape (canola), sunflower, linseed, soybean, and groundnut. Once again, NMR offers a clean, rapid, and accurate alternative to traditional wet chemical techniques and is easier to calibrate than NIR.

Method for seed study

The oil and water in seeds method is based on the inherent capability of NMR to differentiate between signals from solids and bound water, and oil, and then to quantify each separately. MQC benchtop NMRs adhere to the ISO standard method, which is applicable for seeds with up to 10% moisture. Samples with more than 10% moisture can also be easily measured, but these require oven-drying before the NMR analysis.

Advantages of benchtop NMR

NMR is a very stable technique over the long term, requiring minimal recalibration, minimal sample preparation, and no solvent extraction. Because the NMR technique is nondestructive, repeatability measurements can be made conveniently. Sample measurement time is short (typically 20 sec). “Setting-up” samples, with a range of oil and water content, are used to check and maintain the original calibration.

Calibration and results

Figure 3 - Calibration for oil content in rapeseed samples.

Since NMR is a comparative technique, a set of calibration standards of known oil and water content must be obtained before measurements can be made. Thus, the quality of the calibration will always be dependent on the accuracy of the reference data. It is recommended that at least six calibration standards be used with the oil and water contents spread evenly over the range of interest. Real seed samples can be used as calibration standards. These are weighed before and after drying to give their moisture content. The reference values for oil content are normally measured using a solvent extraction technique. Alternatively, a primary oil calibration can be produced using a single sample of the pure oil to be analyzed. Since different types of seed (and oil) result in slightly different NMR signals, better accuracy is achieved when all the standards are of the same species. If measurements on more than one species are required, it is recommended that a separate calibration be created for each.

Figure 4 - Calibration for water content in rapeseed samples.

Nine independently analyzed rapeseed samples were measured. Their oil content varied from 39% to 51% and their water content from 5.2% to 7.1%. Calibrations for oil and water were developed according to ISO 10565 using MultiQuant software, which allows simultaneous calibration and measurement of up to four sample constituents. Measurement time was 16 sec per sample. The resulting calibrations are shown in Figures 3 and 4.

Instrument repeatability for oil was then tested by measuring one sample 10 times without removing it from the instrument. Sample repeatability was tested for oil content by measuring five different portions of the same sample. Instrument repeatability was shown to be 0.03% and sample repeatability 0.21%. The results from both sets of experiments are shown in Table 2.

Recommended NMR instrument configuration

There are two instruments suitable for this application, both of which conform to the industry standard ISO 10565:1998 for a range of sample volumes (given in parentheses): For large sample/seed analysis, the MQC-5 with 51-mm (80-mL) or 40-mm (40-mL) diameter probes is normally used, while for small, low-quantity, or single-seed analysis, the MQC-23 with 26-mm (14-mL), 18-mm (8-mL), or 10-mm (1-mL) diameter probes is preferred.

Table 2 - Instrument and sample repeatability test results for seed samples

The hardware is similar to that described earlier. The EasyCal Oil and Moisture in Seeds application is used with MultiQuant for instrument control. Glass tubes, an installation manual, and a method sheet are also supplied. In addition, an aluminum dry block heater configured for the appropriate tube size, which conditions the samples at 40 °C, is recommended. While the ISO method requires measurement at a nominal room temperature of 17–28 °C, in the authors’ experience, conditioning at 40 °C is preferable when precision measurements are required for oil content only.

Once again, the NMR solution offers multiple advantages over other available options. It operates at high signal sensitivity; requires minimal sample preparation; and an autoweighing facility enhances ease of operation, making operator training simple. Furthermore, operator error is minimized due to automatic postmeasurement calculation of oil content (with respect to 9% moisture and dry weight) while the optional “setting-up” of standards for calibration and maintenance ensures long-term reliability of results.

Conclusion

NMR is an easy-to-use, reliable technique for the quality control of foodstuffs and seeds without the need for solvents with their inherent issues of use and disposal. NMR is recognized by various standards bodies and has been adopted by many laboratories, making it a technique that will remain at the forefront of analytical test facilities.

The authors are with the Oxford Instruments Magnetic Resonance Division, Tubney Woods, Abingdon, Oxon OX13 5QX, U.K.; tel.: +44 (0) 1865 393200; fax: +44 (0) 1865 393427; e-mail: Johnpaul.cerroti@oxinst.com.

Comments