Chapter 8 The Rapid Determination of Fat and Moisture in Foods by Microwave Drying and NMR Analysis Bobbie McManus and Michelle Horn CEM Corporation, Matthews, NC 28106–0200 Abstract Fat and moisture analysis by microwave drying and nuclear magnetic resonance (NMR) is a direct method that yields accurate results in minutes without solvents or the frequent calibration required by indirect methods such as near infrared (NIR) or Fourier transform infrared (FTIR). In the past, NMR was unreliable for determining fat content in foods because the signal from moisture present in the sample interfered with the fat reading. There is now a peer-verified, microwave drying and NMR analysis (AOAC PVM 1:2003) for quantification of the percent- age of moisture and fat in meat products. The procedure involves determining the moisture value of meat samples by microwave drying and using the dried sample to determine the fat value by NMR analysis. Five meat products were analyzed using a CEM SMART System TM (Moisture) and the SMART Trac TM System (Fat). The samples, representing a range of products that meat processors commonly analyze in daily plant operations, included the following: (i) fresh ground beef, high fat; (ii) deboned chicken with skins; (iii) fresh pork, low fat; (iv) all beef hot dogs, and (v) NIST standard reference material samples. The results were com- pared with moisture and fat values derived from AOAC-approved methods, 950.46 Forced Air Oven Drying and 960.39 Soxhlet Ether Extraction. The chapter also discusses the analysis of dairy samples by microwave moisture analysis and NMR fat analysis. These samples include cream, ice cream, and milk with various levels of fat. Results are compared with the Mojonnier and Gerber techniques. Introduction Traditionally, foods have been tested for fat using various wet chemistry methods that are both time consuming and require more than a passing knowledge of chem- istry (1). With the introduction of near infrared (NIR) and Fourier transform infrared (FTIR) instruments, analysis time was reduced to minutes, but the systems required frequent calibration, a time-consuming and expensive process (2). Fat determination by use of a combination of microwave drying and NMR analysis technologies offers the best of both worlds, i.e., the accuracy of traditional wet chemistry methods and the speed of NIR instruments. Copyright © 2004 AOCS Press For many years, laboratories utilized convection or vacuum ovens to dry samples. Although accurate, this was a time-intensive procedure, often requiring 3–72 h to dry a sample completely (2). More recently, halogen and infrared moisture analyzers have reduced drying time to minutes; however, they are not as accurate as traditional oven methods. Microwave moisture analyzers reduce drying time to minutes, while retain- ing the accuracy of oven drying. The SMART System TM (3), as seen in Figure 8.1, incorporates a unique octagonal cavity, which focuses the microwaves on the sample, reducing drying times (Fig. 8.2). It is also the only microwave moisture analyzer that features a temperature-feedback system, which monitors the temperature of the sample and adjusts the power as needed to keep the sample from exceeding the user-defined temperature set point. Additionally, the SMART System (3) continuously monitors weight loss of the sample during drying to ensure accuracy. Microwave drying has become the standard in a variety of industries, and many methods have been written for it by various organizations. The speed and accuracy of microwave drying can greatly help the quality assurance/quality control (QA/QC) of a company and save them money by reducing the amount of out-of-specification product. In the 8 h it takes Fig. 8.1. SMART System microwave moisture analyzer (CEM Corporation). Fig. 8.2. Patented SMART System octagonal cavity efficiently focus- es microwave energy to decrease drying times. Copyright © 2004 AOCS Press to conduct a single moisture analysis in a standard drying oven, a company can con- duct over 120 analyses in a microwave moisture analyzer. Nuclear magnetic resonance (NMR) imaging is nondestructive and noninvasive, using a magnet and an electromagnetic coil to force the hydrogen atoms in a substance to align with the magnetic field (Figs. 8.3 and 8.4). It then sends a pulse of radio-fre- quency energy through the field, briefly creating a stronger field with which the hydro- gen atoms align (Fig. 8.5). When the pulse ends, the atoms “relax” back to their origi- nal alignment and emit a weak electrical signal known as transverse relaxation (T 2 ) decay or free induction decay (FID), which is then measured by the NMR (Fig. 8.6). Fig. 8.3. A sample is placed in the magnetic field. Fig. 8.5. An RF pulse forces atoms to align temporarily. Fig. 8.4. Hydrogen atoms align to the magnetic field. Fig. 8.6. The atoms “relax” and the instrument interprets the signal. Copyright © 2004 AOCS Press The hydrogen atoms in proteins and carbohydrates relax faster than those in fat; thus, the instrument can distinguish between the signals to determine the amount of fat in a substance (4,5). The term “nuclear” in NMR refers to the instrument's ability to analyze the nuclei of a sample. (The instrument does not generate any ionizing radiation.) NMR can cause many nuclei to generate a signal, but most research has been performed on the detection of signals from the 1 H nucleus, a branch of science commonly known as “proton NMR.” After a food sample is dried with microwave energy, the remaining components that contain significant amounts of protons are fats, proteins, and carbohy- drates, all of which have different transverse relaxation times. The relaxation times for fats are significantly longer than those of proteins and carbohydrates. Fats typically have transverse relaxation times of <10 µs or more, whereas proteins and carbohy- drates demonstrate “solid-like” behaviors and generally have relaxation times ≤10 µs. Thus, the NMR is able to distinguish between the signals of different components; after exciting the molecules and letting the “solid-like” signals of protein and carbohy- drates decay, the remaining signal from the dried sample will be fat. In the past, NMR was not utilized for determining fat content in many foods because the signal from moisture present in the sample interfered with the fat reading. However, by first drying samples using microwave energy, an accurate NMR analysis can be performed. The SMART Trac TM System is a synergy of advanced microwave drying and NMR technologies that culminates in a truly unique fat analyzer that yields accurate results in minutes (Fig. 8.7). The SMART Trac is designed to rapidly mea- sure moisture and fat content in all types of foods in <5 min without solvents, chemi- cals, or time-intensive maintenance. The system takes a direct measurement of total fat, including both free and chemically bound fat in solids, liquids, and slurries. The system is applicable to a full range of food products from low-moisture samples, including powders and snack foods, to high-moisture samples, such as meat and dairy products. Unlike wet chemistry methods, the SMART Trac is easy to run and does not require a knowledge of chemistry. The system measures fat throughout the entire sam- ple and is not affected by variability in the product (e.g., changes in color, texture, or composition), whereas NIR and FTIR systems measure fat in only a small portion of the sample and must be recalibrated when there are changes in the product. Fig. 8.7. SMART Trac fat and moisture analyzer (CEM Corporation). Copyright © 2004 AOCS Press Experimentation After first discussing the equipment, this chapter will describe the methodology for analysis of a food sample in general and then discuss the specifics of meat and dairy samples. This new technology for fat analysis has generated a great deal of interest and there is now a peer-verified, microwave drying and NMR analysis method (AOAC PVM 1:2003) for the identification of the percentage of moisture and fat in meat products. The procedure involves determining the moisture value of meat samples by microwave drying and using the dried sample to determine the fat value by NMR analysis. The samples, which represent a range of products that meat processors commonly analyze in daily plant operations, included the follow- ing: (i) fresh ground beef, high fat; (ii) deboned chicken with skins; (iii) fresh pork, low fat; (iv) all beef hot dogs; and (v) NIST standard reference material samples. The results were compared with moisture and fat values derived from AOAC- approved methods, 950.46 Forced Air Oven Drying and 960.39 Soxhlet Ether Extraction (2). A method for dairy samples is currently being evaluated by the AOAC, and approval is expected by mid-2004. Equipment The microwave drying/NMR method for this experiment uses a SMART Trac System comprised of a SMART System TM microwave moisture analyzer and a SMART Trac low-resolution time domain NMR (LR-NMR). Microwave Moisture Analyzer. The SMART System moisture analyzer features 0.2 mg H 2 O sensitivity, a moisture range of 0.01–99.99% in liquids, solids, and slurries, 0.01% resolution, and an automatic electronic balance (0.1-mg readabili- ty). The system has built-in power management and temperature feedback control for reproducible results. Optimization of power flow into the unique octagonal cav- ity provides the fastest possible drying times and ensures that high-carbohydrate samples will not overheat. The system can store 100 methods and up to 300 test results, as well as perform a statistical analysis of the process including an average of results, SD, minimum/maximum range, and graphing in reference to target val- ues. NMR Processor and Magnet. The SMART Trac NMR Radio Frequency Pulse Generator uses a pulse power of 250 W nominal, variable pulse times in 100-ns increments, transmission and reception phases selectable at 0, 90, 180 and 270°C; and, nominal 90° pulse time of 4 µs (18-mm probe). The magnet is permanent and thermally stabilized, 0.47 T (20 MHz); homogeneity is better than 10 ppm. Signal detection is accomplished by dual channel (quadrature) detection with programma- ble low-pass filtering and a programmable data acquisition rate up to 4 MHz per pair of points. Figures 8.8–8.10 demonstrate the general method of processing a sample. Copyright © 2004 AOCS Press Meat Applications Fat is an important component in the meat industry. A less expensive ingredient than meat, fat adds flavor, but too much of it renders a product extremely difficult to market in an ever increasingly health conscious society. The SMART Trac enables manufacturers of meat and meat products to maximize yield and meet customer expectations, while simultaneously helping to control production costs. Fig. 8.8. A sample is spread or absorbed onto a proton-free medium, typically a glass fiber sample pad, and then continu- ously weighed while it is dried in the microwave system at a controlled temper- ature. It is then cooled to 45°C. Total time for drying and cooling is 2–3 min. Fig. 8.10. Because the sample is already conditioned at a controlled temperature, the NMR analysis can be immediately performed, typically requiring <1 min. Fig. 8.9. The sample is then placed on a proton-free film (Trac film), rolled up, and pressed into a Teflon tube for the NMR analysis. Copyright © 2004 AOCS Press Preparation of Meat Samples According to an AOAC-Approved Peer- Verified Method for the Rapid Determination of Fat and Moisture To prevent water loss during preparation and subsequent handling, do not use small samples. Keep ground material in glass or similar containers with air and watertight covers. Prepare samples for analysis as follows: 1. Collect ~5 pounds of freshly processed product and store at ≤4°C until analyzed. 2. Dice nonground meat product samples into ~2-inch (5.08 cm) cubes and pass rapidly 2 times through a Hobart grinder (model 4612) or equivalent equipped with a 3/16 inch (0.1875 inch or 0.4763 cm) plate. 3. Homogenize ground meat products in a Robot Coupe bowl chopper (model R6 or equivalent) to a paste or paté consistency. Place ground material in a chilled bowl chopper (4°C) and chop for 30 s; scrape down the inner walls and bot- tom of the bowl with a spatula (plastic or rubber spatula with ~2 × 4 inch straight-edge blade) and add the material gathered to the body of the test sam- ple. Repeat the process for an additional 30 s. 4. After homogenization, collect samples and store in plastic containers with screw cap lids. If the samples are not to be analyzed immediately, store them at ≤4°C until needed. Procedure for Fat Determination in a Meat Sample 1. On the SMART System Main Menu screen, select Load Method, then select the appropriate preprogrammed item to be analyzed, (e.g., ground beef). Note: Different types of sample matrices and fat will exhibit different responses on the NMR system. To obtain accurate fat readings, two or more samples of the specif- ic sample type must be analyzed by the AOAC method. The samples should cover the entire fat range to be analyzed. Preferably, one high-fat reference sam- ple and one low-fat reference sample should be analyzed. The reference values are typed into the SMART Trac system and then replicate runs of each sample are performed to determine the appropriate NMR signal values for that specific sample type. After completing the reference scans, the SMART Trac system will establish a linear relationship for fat determination for that type of sample. 2. Press the Ready Key to initiate the analysis. Place two glass fiber sample pads (square) in the SMART System Moisture/Solids Analyzer microwave chamber on the balance and press Tare on the keypad. Tare weight will be automatically recorded. 3. With a Teflon-coated spatula, transfer ~3–5 g of sample, from the center of the refrigerated sample container to the center of one of the tared sample pads. Spread the meat sample evenly across the square pad (see Figs. 8.11 and 8.12). 4. Cover the sample with the other tared square pad as if making a sandwich and place the pads on the balance in the SMART System Moisture Analyzer. Copyright © 2004 AOCS Press 5. Dry sample by pressing Start on the keypad. A temperature feedback system allows rapid temperature measurement of the sample during drying to adjust the microwave power delivery. Percent Moisture will be displayed on the screen (± 0.01%) after the sample has dried to a constant weight. Note: Five short beeps will be heard when drying is complete. 6. Remove pads and roll both in Trac Film (see Figs. 8.13 and 8.14). 7. Compress the rolled sample in the plastic sleeve using the compression tool, insert sample into the NMR chamber for analysis. [The sample is placed in the core of an 89-kg magnet and pulsed with radio frequency (RF) energy while in the static magnetic field. The resulting signal is recorded and analyzed for the total proton activity of fat present in the sample. Proprietary software analyzes the NMR data and provides the moisture and fat results.] 8. Press Ready to continue the fat analysis, then press Start to analyze for fat. Percent Fat will be displayed on the screen (± 0.01%). Fig. 8.11. If the sample is in a paste, semisolid, or crumb form or a raw or skeletal meat product such as fresh pork, ground beef, or chicken, place the sample on the end of a spatula and spread it across one end of the pad. Then spread the sample to a uniform thickness covering ~90% of the pad surface area. Fig. 8.12. If the sample contains bound water such as an all-meat emul- sion, cooked all-meat sausage, sausage with extenders, semidry sausage, or ham, place the sample on the end of the spatula and apply the sample to the middle of the pad. Then spread the sample around the pad in a circle. Copyright © 2004 AOCS Press Fat and moisture determinations on several different meat products and NIST standards were performed in laboratories at Texas A&M (TAM) and CEM Corporation. The results for beef, all beef hot dogs, and chicken are presented in Tables 8.1, 8.2, and 8.3, respectively. Fig. 8.13. Place the two square pads and dried sample in the center of the Trac film. Fold the left corner of the film and pads as illustrated. Fold the right corner. Pull the lower edge of the film and sample pads toward the top and begin to roll them into a tube. Fig. 8.14. For samples that are rigid after being dried and more difficult to roll into a cylinder, prepare the pads as illustrated. Copyright © 2004 AOCS Press Dairy Applications Fat and moisture/solids contents are important components in the dairy industry. Cream and milk are bought and sold on the basis of the amount of fat they contain. Because fat is an expensive element in dairy products, it is essential that companies are certain of the fat content of the cream obtained from their suppliers and that they tightly control the fat in their products for quality and consistency, cost man- agement, and regulatory requirements. Similarly, effective moisture/solids control maximizes yield. Generally, dairies benefit from controlling solids content in all of their products because moisture, of course, is a less expensive ingredient than milk solids. Thus, fat and moisture/ TABLE 8.1 AOAC Results from TAM (AOAC Method 950.46 and 960.39) for Fresh Ground, High-Fat Beef Weight Moisture Weight Fat Sample ID (g) (%) (g) (%) 1 4.9750 40.44 3.6150 46.03 2 3.4373 40.42 3.7455 45.86 3 3.7334 40.35 3.8465 45.83 4 3.6018 40.26 3.3339 45.54 5 3.6140 40.49 3.3647 45.92 6 4.2714 40.34 3.6816 45.76 7 3.3022 40.41 3.8443 45.64 8 3.4890 40.44 2.2901 45.92 9 3.2051 40.30 2.2819 45.93 10 3.3612 40.45 2.2736 45.99 Mean 40.39 45.84 SD 0.074 0.157 SMART Trac results from CEM SMART Trac results from TAM Microwave NMR Microwave NMR Weight moisture fat Weight moisture fat Sample ID (g) (%) (%) (g) (%) (%) 1 3.6893 40.19 46.34 3.3149 40.69 45.68 2 3.5562 40.13 46.15 3.8042 40.23 46.25 3 4.1158 40.26 46.08 3.5233 40.18 46.27 4 3.9324 40.20 46.25 3.5591 40.68 45.93 5 4.0714 40.19 46.19 3.9022 40.24 46.24 6 3.8083 40.35 45.99 3.7258 40.32 46.40 7 3.7083 40.44 45.59 3.2478 40.42 45.94 8 3.4759 40.33 45.65 2.9598 40.39 45.92 9 3.4957 40.36 45.50 3.3260 40.42 45.85 10 3.6916 40.32 45.87 3.4150 40.59 45.54 Mean 40.28 45.96 40.42 46.00 SD 0.098 0.295 0.184 0.280 Copyright © 2004 AOCS Press [...]... 2004 AOCS Press TABLE 8. 3 AOAC Results from TAM (AOAC Method 950.46 and 960.39) for Fresh Chicken with Skin Sample ID 1 2 3 4 5 6 7 8 9 10 Mean SD Weight (g) Moisture (%) Weight (g) Fat (%) 5 .88 88 5.9199 6. 685 1 5.9594 6.2424 6.5156 6.6359 6.2113 6.4292 6.4023 74.52 74.56 74.59 74.55 74.52 74.60 74.60 74. 58 74.64 74.56 74.57 0.0 38 3.3 288 3.4672 3.5339 3.6244 3. 487 7 3.7604 3.37 48 3.5340 3.6059 4.2455... Sample ID 1 2 3 4 5 6 7 8 9 10 Mean SD Weight (g) 4.0 588 4.3006 3.5390 4.0012 3.65 98 3.9 586 4.1737 3.6191 3. 587 8 3.7391 Microwave moisture (%) NMR fat (%) 51.55 51.42 51.64 51.53 51.46 51.47 51. 38 51.26 51.44 51.17 51.43 0.1 38 30.63 30.64 30. 28 30.46 30.65 30.36 30.72 30.57 30.42 30.71 30.54 0.154 Fat (%) 30.59 30.61 30.44 30.53 30.63 30.36 30.72 30.20 30.54 30. 18 30. 48 0. 183 SMART Trac results from...TABLE 8. 2 AOAC Results from TAM (AOAC Method 950.46 and 960.39) for All Beef Hot Dogs Sample ID 1 2 3 4 5 6 7 8 9 10 Mean SD Weight (g) Moisture (%) Weight (g) 2.4646 2.2422 2.0909 2.0910 2.1961 2.40 98 2.2961 2. 285 9 2.4673 2. 081 6 51.51 51.45 51.64 51.54 51.63 51.75 51 .87 51 .89 51 .80 51.69 51. 68 0.152 3.2652 3.6297 4.1140 3.1469 3.5239 3.6645 3.4951 3.13 08 3.2411 3.1027 SMART Trac... 7. 08 7.06 7.16 7. 08 7.41 7.20 7.24 0. 186 SMART Trac results from CEM Sample ID 1 2 3 4 5 6 7 8 9 10 Mean SD Weight (g) 4.0309 3. 489 1 3.7672 3.4990 3 .89 61 3.74 38 3.7 383 3.7661 3 .89 05 4.1224 Microwave moisture (%) NMR fat (%) 74.55 74.51 74.66 74.65 74.53 74.42 74.65 74.25 74.25 74.31 74. 48 0.162 7.25 7.34 7.33 7.32 7.29 7.26 7.24 7.25 7.30 7.27 7.29 0.036 SMART Trac results from TAM Weight (g) 2.73 48. .. 0.37 4.09 1. 48 10.06 9.62 4.15 3.52 0.35 1 .88 0.05 4.06 3.01 1 .89 0.04 4.04 3.01 Cheddar cheese Reduced-fat cheddar cheese Nacho cheese 33.2 20 .8 8.1 33.19 20 .85 8. 05 Sour cream Light sour cream Fat-free sour cream 17.35 8. 46 2.24 17.4 8. 52 2.21 Copyright © 2004 AOCS Press Fig 8. 15 Spreading technique for liquid samples 2 Press the Ready Key to initiate the analysis Place two glass fiber sample pads... 2. 681 5 3.2946 3.01 28 3.6523 3.35 48 4.1013 2 .84 25 3.3257 Microwave moisture (%) NMR fat (%) 74. 28 74.23 74. 38 74. 18 74.32 74.42 74.55 74.40 74.41 74.54 74.37 0.121 7.29 7.14 7.22 7.21 7. 08 7.16 7.07 7.10 7.07 7.11 7.15 0.074 1 On the SMART System Main Menu screen, select Load Method, then select the appropriate preprogrammed item to be analyzed, (e.g., cream) Note: Different types of sample matrices and. .. Remove pads and roll both in Trac Film (see Figs 8. 13 and 8. 14 in the previous section) 7 Compress the rolled sample in the plastic sleeve using the compression tool, and insert the sample into the NMR chamber for analysis [The sample is placed in the core of an 89 -kg magnet and pulsed with radio frequency (RF) energy while in the static magnetic field The resulting signal is recorded and analyzed... AOAC International, Gaithersburg, MD 3 Keeton, J., Moser, C., and Haire, B., in press, AOAC Peer Review Study: The Rapid Determination of Moisture and Fat in Meats and Dairy Products by Microwave Drying and NMR Analysis, Official Methods of Analysis of AOAC International Updates, AOAC International, Gaithersburg, MD 4 Macomber, R.S (19 98) A Complete Introduction to Modern NMR Spectroscopy, John Wiley... activity of fat present in the sample Proprietary software analyzes the NMR data and provides the moisture and fat results.] 8 Press Ready to continue the fat analysis, then press Start to analyze for fat Percent Fat will be displayed on the screen (± 0.01%) Summary Fat and moisture determination by microwave drying/NMR analysis is a direct method that provides accurate results in minutes without the... 1.9061 2.9303 3.2794 3.4650 3.3049 2 .85 97 3.0330 2.7 080 2.7 289 Microwave moisture (%) NMR fat (%) 51.79 51.96 51. 58 51.59 51.97 51.47 51.66 51.97 51.69 51.49 51.72 0.195 30.30 30.77 30.60 30.29 30.36 30.63 30.45 30.40 30.63 30.71 30.51 0.175 solids analyses are very beneficial in the production of cheese, milk, sour cream, ice cream, butter, yogurt, cottage cheese, cream, and other dairy products The ability . 3.5239 30.63 6 2.40 98 51.75 3.6645 30.36 7 2.2961 51 .87 3.4951 30.72 8 2. 285 9 51 .89 3.13 08 30.20 9 2.4673 51 .80 3.2411 30.54 10 2. 081 6 51.69 3.1027 30. 18 Mean 51. 68 30. 48 SD 0.152 0. 183 SMART Trac. 7.25 2.73 48 74. 28 7.29 2 3. 489 1 74.51 7.34 3.4454 74.23 7.14 3 3.7672 74.66 7.33 2. 681 5 74. 38 7.22 4 3.4990 74.65 7.32 3.2946 74. 18 7.21 5 3 .89 61 74.53 7.29 3.01 28 74.32 7. 08 6 3.74 38 74.42 7.26. 45 .86 3 3.7334 40.35 3 .84 65 45 .83 4 3.60 18 40.26 3.3339 45.54 5 3.6140 40.49 3.3647 45.92 6 4.2714 40.34 3. 681 6 45.76 7 3.3022 40.41 3 .84 43 45.64 8 3. 489 0 40.44 2.2901 45.92 9 3.2051 40.30 2. 281 9