- Email: [email protected]
Radio frequency cooking of ground, comminuted and muscle meat products
Meat Science 65 (2003) 959–965 www.elsevier.com/locate/meatsci
Radio frequency cooking of ground, comminuted and muscle meat products L. Laycocka, P. Piyasenab, G.S. Mittala,* a School of Engineering, University of Guelph, Guelph, Ontario, Canada N1G 2W1 Food Research Program, Agric & Agri-Food Canada, Guelph, Ontario, Canada N1G 2W1
b
Received 25 September 2002; received in revised form 12 November 2002; accepted 12 November 2002
Abstract The effect of radio frequency (RF) cooking, on the quality (colour, water holding capacity, texture), heating rate, and temperature history of three types of meat products (ground, comminuted and muscle), was investigated after heating to centre temperature of 72 C in a 1.5 kW RF heater operating at 27.12 MHz. The results were compared with that obtained from heating in a water bath. RF cooking of processed meat products resulted in a decreased cooking time (5.83, 13.5, and 13.25 min for ground beef, comminuted meat, and muscle, respectively compared to 151, 130, and 109 min in water bath), lower juice losses, acceptable colour, water holding capacity and texture. The results indicate that when using RF, ground beef has the highest power efficiency (60.17%) followed by comminuted meat (44.41%), and muscle (43.38%). However, the texture of ground beef was too chewy and elastic. The muscle’s colour was inferior. The comminuted and muscle meat products exhibited average energy efficiency with improved texture. The comminuted meat displayed a large cross-sectional temperature differential, possible due to uneven salt distribution. The well mixed comminuted and ground meat products appeared to be most promising for RF cooking. # 2003 Elsevier Ltd. All rights reserved. Keywords: Processed meat; Meat cooking; Meat heating; Meat quality; Meat colour; Meat texture; Texture profile analysis
1. Introduction The safety of proper cooking of meat and meat products is of increasing concern. Conventional food heating methods require that food be heated externally through conduction, convection or radiation. The main disadvantages of conventional methods are longer cooking time and non-uniform heating. The cooking time is lengthy for conventional methods as compared to dielectric heating such as radio frequency (RF) or microwave. With RF, due to volumetric heating, the product is heated internally, and not through the surfaces. The product forms a dielectric between the two electrodes acting as capacitor plates that are alternatively charged from positive to negative several millions times a second (27 MHz). The polar molecules in the product try to realign themselves with the polarity of the electrodes, causing a great deal of internal friction * Corresponding author. Tel. +1-519-824-4120x2431; fax: +1-519836-0227. E-mail address: [email protected] (G.S. Mittal). 0309-1740/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0309-1740(02)00311-X
to occur by which the product is heated. Also, RF cooking generally occurs uniformly across the sample and prevents the occurrence of degradation at the outside surfaces (Zhao, Flugstad, Kolbe, Park, & Wells, 2000). RF heating is currently practiced in the baking industry to remove excess moisture during post baking. Thawing of meats is also a common practice using RF technology (Anon., 1996; McCormick, 1988). Cooking and/or pasteurization of only a few selected meat products have been investigated. Houben, Schoenmakers, van Putten, van Roon, and Krol (1991) reported RF heating to be the most promising technique for the pasteurization of moving sausage emulsions in terms of penetration depth, energy efficiency and product quality. The pasteurization of cured hams was investigated by Bengtsson and Green (1970) and compared with conventional hot water treatments. RF treatment at 60 MHz reduced the processing time by one third, reduced juice losses and improved sensory quality of the meat. The only disadvantage was the increased bacterial count after three weeks of storage. Orsat, Bal, and Raghavan
960
L. Laycock et al. / Meat Science 65 (2003) 959–965
(1999) also used RF to pasteurize ham and found that the shelf life can be improved by using RF energy. Procter Strayfield at Horsham, PA developed a ‘‘Magnatube’’ pasteurization system capable of cooking scrambled eggs and skinless meatloaf (Zhao et al., 2000). Uniform temperature distribution and a high yield were achieved. The objective of this study was to evaluate the influence of RF cooking on the quality of each of the three types of meat products (ground, comminuted and muscle) after being heated to a centre temperature of 72 C.
2. Materials and methods 2.1. Equipment The experiment was performed at a constant frequency of 27.12 MHz with a heater of 1.5 kW (Strayfield Ltd, Reading, UK). The heater was comprised of two components: a RF generator and an applicator circuit (Awuah, Ramaswamy, & Piyasena, 2002; Strayfield, 1993). The applicator consisted of two electrodes, tuning capacitor plates, an inductance coil and an applicator cylinder. The cylinder was made of polytetrafluoroethylene (PFTE), which is RF inert. The cylinder (ID=0.0952 m, OD=0.121 m, height from inside=0.203 m, height from outside=0.215 m), with the product inside, was set between two electrodes. The electrodes wrapped around the applicator cylinder allowing a very small gap to remain between them. The electrodes were parallelograms and formed to wrap around the applicator cylinder such that every portion of the sample was subjected to dielectric heating. The electrodes had a diameter of 0.125 m and a height of 0.1496 m. The power from the RF generator was adjusted by changing the distance between the tuning plates of the tuning capacitor. If the plates are brought close together, the RF power applied to the product will increase and vise versa. This was accomplished manually by using a gear mechanism outside the RF oven (Strayfield, 1993). The active current passing through the product was maintained at constant level. When the product began to cook, the meat properties changed thus changing the active current. Active currents of 425 mA for ground beef, 320 mA for comminuted meat and 330 mA for whole muscle, were maintained by using the gear mechanism. The standing current, the current needed to maintain oscillations of the RF field when the applicator cylinder is empty, was 175 mA. The actual current is the difference between the two, resulting in 250, 145 and 155 mA. The power input to the RF heater is the result of the actual current multiplied by the high voltage. The high voltage for this unit was approximately 4.8 kV (Strayfield, 1993) resulting in a power input (Ps)
of 1200, 696 and 744 W, respectively, for ground beef, comminuted meat and whole muscle in this study. 2.2. Meat products Three types of meat products were prepared; ground beef, 2% salt comminuted beef and beef muscle. The ground beef used for both the ground and comminuted samples, was obtained from a local meat plant. The beef muscle product was obtained from Rowe Farm Meats, Guelph Ontario, Canada. The comminuted sample was prepared by evenly spreading 2% NaCl by mass around the sample then passing the product through a grinder (Hobart, Toronto, Canada; model #4532) with a 10 mm dia. openings twice. The muscle was cut into cylindrical pieces that would easily pack into the RF applicator cylinder. All raw samples had the same dimensions: a diameter of 0.0952 m and a height of 0.1496 m. The ground and comminuted meats were initially mixed thoroughly for 1 min with an electric mixer. In a 4 C refrigerator, the samples were layered into the custom made applicator cylinder, followed by a firm packing action after each layer to remove any air pocket that may have formed. The raw ground beef sample (1.15 0.0467 kg) contained 56% moisture, 27% protein, 16% fat and 1% ash. The raw comminuted sample (1.11 0.01 kg) contained 60% moisture, 24% protein, 13% fat, and 3% ash. The raw muscle sample (1.145 0.015 kg) contained 72% moisture, 22% protein, 5% fat and 1% ash. 2.3. Cooking and temperature measurement Two fiber-optic temperature probes (model # SFW-2, Luxtron Corporation, Santa Clara, CA) were calibrated using an oil bath. The sensors were connected to a signal converter and a display unit (755 multi-channel fluoroptic thermometer, Luxtron Corporation, Santa Clara, CA). One probe recorded the centre temperature (75 mm from the top surface) while the other measured the temperature 10–15 mm from the side (75 mm from the top surface). The temperature was manually recorded every 15 s until the centre temperature reached 72 C at which point the RF generator was ceased. The juice surrounding the product was removed and the product was weighed to determine the juice loss. The samples were then refrigerated at 4 C overnight. Three replications of each product were conducted. These products were also cooked in a water bath (VWR International, Sheldon Manufacturing, Cornelius, OR; model #1217). The dimensions of these samples were similar with a 9.5 cm diameter and 12 cm height. They were packed inside a plastic bag to prevent leakage and a sheet of aluminum foil to maintain their shape. The height of the water was about 12 cm inside the water bath, and the aluminum foil kept open at the
L. Laycock et al. / Meat Science 65 (2003) 959–965
top for the insertion of the temperature probe. One temperature probe (Omega, Laval, Canada; model HH23, microprocessor thermometer, type T thermocouple) was inserted in the product centre. Initially, the temperature of the water bath was 65 C and was gradually increased (0.2 C/min) to a final temperature of 85 C. The temperature was recorded every few minutes to obtain a temperature-time history. The product was removed from the water bath when the center temperature reached 72 C.
961
#5.0.1 was used to record the data. The results were then analyzed using a Microsoft Excel spreadsheet to compare the RF results to those measured using the water bath. 2.7. Data analysis All treatments were carried out three times. Duncan analysis for ranking means was conducted using the Statistical Analysis System (SAS, 1996).
2.4. Colour 3. Results and discussion The colour test measured the surface colour of the cooked meat with respect to lightness (L*), redness (a*) and yellowness (b*). For meat products, L* and a* are the most important parameters. The lighter and redder the product, the better the quality. A Minolta colorimeter (model CR-310, Minolta Co., Ramsey, NJ) and data processor were first calibrated with a known light source. Three colour replications of each cooked crosssection were then taken. 2.5. Water holding capacity (WHC) The centrifugal method of the water holding capacity (WHC) test measured the ability of the meat to absorb water. A product with a greater WHC is juicier and is generally preferred. Three replications of 5 g samples of each cooked product were placed inside centrifuge tubes while 8 ml of water was added to each tube. The tubes containing samples were incubated at 20 C for 1 h while the rotor was cooled in a refrigerator to 5 C. The centrifugation occurred at 7500 rpm for 7.5 min at 20 C (Zhang, Mittal, & Barbut, 1995). The WHC was then calculated using:
3.1. Cooking time The ground beef samples were cooked in 5.83 0.417 min using RF. The comminuted product was uniformly cooked in an average time of 13.5 2.0 min. The muscle was cooked in 13.25 1.75 min. The samples cooked in the water bath (WB) took much longer. The cooking times for the ground, comminuted and muscle samples were 151 1, 130 2 and 109 2 min, respectively. Cooking time in WB could be reduced by using faster rate of water temperature increase or higher initial water temperature. Figs. 1–3 compare the cooking times for ground, comminuted and muscle meat, respectively. The time taken to cook a 1.15 kg meat sample ranges from 5 to 15 min using RF energy compared to WB processing times of 107–130 min. Thus same amount of meat can be cooked approximately six times faster using RF. 3.2. Energy consumption and efficiency The power requirement of the process was calculated by determining the power required by the product over
WHC ¼ ðsolution added to the sample ðgÞ solution removed after centrifugation ðgÞÞ 100=ðmeat sample mass ðgÞÞ 2.6. Textural profile analysis (TPA) Texture profile analysis (TPA) measured the hardness, cohesiveness, springiness, gumminess and chewiness of a product. A universal testing machine (Instron, Burlington, Canada; model # 4204) was used with a 50 kN load cell. All samples were tested at room temperature having a diameter (D) of 25.5 mm and height (L) of 17 mm, resulting in a D/L ratio of 1.5. The samples were compressed down 12.75 mm or 75% (Mittal, Nadulsky, Barbut, & Negi, 1992) at a crosshead speed of 100 mm/ min. The test ran for the duration of two compression cycles. Labview data acquisition software, specifically custom-made interface software called Adlab version
Fig. 1. Temperature–time history of cooking ground beef (GB) using RF and a hot water bath (WB).
962
L. Laycock et al. / Meat Science 65 (2003) 959–965 Table 1 Mean colour parameters of both RF and water bath (WB) cooked meat products Product type
Cooking medium
Mean lightness (L*)
Mean redness (a*)
Ground beef Ground beef
RF WB
44.261.10a 43.191.45a
8.22 0.93a 8.85 0.30a
Comminuted Comminuted
RF WB
42.791.06b 41.420.44b
7.33 1.13b 7.93 1.22b
Muscle Muscle
RF WB
49.831.86c 52.500.71c
7.35 0.62b 7.96 0.14b
Means with the same letter in a column are not different at 95% level.
Fig. 2. Temperature–time history of cooking comminuted (COM) beef using RF and a hot water bath (WB).
Fig. 3. Temperature–time history of cooking beef muscle using RF and a hot water bath (WB).
the entire cooking time. The power required can be calculated using: Pw ¼ mcp T=t where Pw is the power required, W, cp is the specific heat, J/(kg C), m is the mass, kg, T is the temperature change, C, and t is the cooking time, s. The overall power efficiency can be calculated using (Houben et al., 1991): Pefficiency ¼
Pw 100 Ps
where Pw is the power required by the product, and Ps is the power supplied to the RF generator. The results indicate that when using RF, ground beef has the highest power efficiency of 60.17 2.26%. Comminuted meat has the next highest power efficiency
of 44.41 6.27%. Muscle product resulted in a power efficiency of 43.38 5.17%. Bengstsson and Green (1970) pasteurized ham with RF energy and reported the highest power efficiency of 25%. It was noted that the requirement for an industrial application would have to be 50–60% efficiency. Houben et al. (1991) performed pasteurization of moving sausage emulsions and reported average power efficiencies of 25–30%, which could be improved by specifically designing the generator-heating unit combination. Zhao et al. (2000) suggested that the energy losses within an RF system include matching networks, couplers and the capacitive electrode system. Therefore, overall energy efficiencies would probably not exceed 50–60% with the present technologies available. The results of this study indicate that RF energy is much more energy efficient as compared to WB processing. The energy efficiency of ground beef using RF was 48.13 1.81% as compared to WB heating results of 2.53 0.06%. The comminuted meat cooked with RF had an energy efficiency of 20.61 2.91% as compared to the WB cooked samples of 2.53 0.06%. The muscle RF cooked samples had an energy efficiency of 21.52 2.57% as compared to the WB samples of 2.61 0.07%. The low efficiency in case of WB cooking was due to the cooking of small amount of meat product in a large amount of water and large losses of heat to environment. Hence this is not the true efficiency achieved at industrial scale. Nevertheless, it is much more efficient to use RF as opposed to WB processing to cook all three meat products. Altering the design of the generator or creating a continuous flow tube could improve the energy efficiencies of both RF cooked comminuted and muscle meats. 3.3. Colour The results from the colour test are summarized in Table 1. In terms of L* and a*, the RF cooked meat products samples were similar to those WB cooked. The differences were so small that these were also not visible to the naked eye.
963
L. Laycock et al. / Meat Science 65 (2003) 959–965
3.4. Water holding capacity (WHC) Table 2 summarizes the results of water holding capacity using the centrifugal method. RF comminuted meat did not absorb as much water as compared to WB cooking. This may cause the meat to be little drier to the palate but may be acceptable. The RF muscle samples have a much better WHC as compared to the WB cooked samples. In terms of WHC, the results do not suggest that RF energy cause the meat to become less absorbent and in turn less juicy. Juice loss during cooking is another method of determining the juiciness of a product. Table 3 summarizes
the juice loss percentage of both RF and WB cooked products. The RF cooked samples seemed to hang on to their moisture during cooking. This would have a lot to do with cooking time, the longer the cooking time, the drier a product becomes. These results suggest that RF heating reduces moisture loss and the result is a juicier product. This finding has also been reported earlier (Anon., 1996; Bengtsson & Green, 1970). Moisture content was similar for both cooking methods (WB and RF) suggesting that the WB cooked products have a tendency to lose more fat than water during cooking, keeping the water content at the same level. 3.5. Texture profile analysis
Table 2 Mean water holding capacity (WHC) values of both RF and water bath (WB) cooked meat products Product type
Cooking medium
Water holding capacity
Ground beef Ground beef
RF WB
6.630.09b 6.610.05c
Comminuted Comminuted
RF WB
6.630.02b 6.670.06a
Muscle Muscle
RF WB
6.610.08c 6.550.06d
Means with the same letter in a column are not different at 95% level. Table 3 Mean juice loss percentages of both RF and water bath (WB) cooked meat products Product type
Cooking medium
Juice loss (%)
Ground beef Ground beef
RF WB
25.5 3.7bc 26.0 0.1b
Comminuted Comminuted
RF WB
20.9 1.2d 23.4 0.6c
Muscle Muscle
RF WB
28.7 6.9ab 31.2 0.3a
Means with the same letter in a column are not different at 95% level.
The TPA parameters of ground beef, comminuted meat, and muscle cooked with RF and WB are shown in Table 4. 3.5.1. Ground beef The mean hardnesses (1 and 2), cohesiveness and gumminess were similar for RF and WB cooked ground beef. The springiness represents the elastic properties of the meat. The higher the springiness, the more elastic or rubbery characteristics the product will exhibit. The ground beef cooked with RF had significantly higher springiness (9.85 0.56 mm) than WB cooked (4.91 0.55 mm). Thus, the RF samples were more elastic, and may contribute to the product chewiness. A large difference was observed with respect to chewiness. The chewiness value of the RF cooked ground beef (926.8 171.9 mJ) was significantly higher compared to WB cooked (430.1 103.7 mJ). The product cooked with RF was much too chewy. Thus, RF cooked ground beef was significantly more elastic and chewy. Van Roon, Houben, Koolmees, Krol, and van Vliet (1994) also reported that meat dough heated by RF was much firmer than those WB heated. The potential for RF cooking of ground beef looks promising, however, slightly slower heating is required and better process conditions are needed for the integration of such a process in industrial practices.
Table 4 Mean and standard deviation of TPA parameters of RF and WB cooked ground, comminuted and muscle beef products Meat product
Cooking medium
Hardness1 (N)
Hardness2 (N)
Cohesiveness
Springiness (mm)
Ground Ground
RF WB
313.337.2b 32934.3b
225.831.0b 238.428.5b
0.290.02b 0.270.02b
9.850.56a 4.910.55d
926.8171.9b 436.1103.7c
93.814.7 87.813.3b
Comminuted Comminuted
RF WB
206.515.6c 302.235.4b
152.514.9c 233.933.0b
0.270.02b 0.280.01b
8.160.63b 5.760.43d
459.991.0c 497.679.8c
568.0c 86.312.0b
Muscle Muscle
RF WB
709.2143.8a 698.2126.8a
588.1122.8a 592.6118.3a
0.430.04a 0.390.06a
6.60.68cd 7.260.28c
2059.1531.3a 2001523.9a
311 70.0a 275.770.1a
Means with the same letter in a column are not different at 95% level.
Chewiness (mJ)
Gumminess (N)
964
L. Laycock et al. / Meat Science 65 (2003) 959–965
3.5.2. Comminuted beef Average values of cohesiveness and chewiness were similar for RF and WB cooked comminuted beef. If the hardness value is too high, the product is too tough. If the hardness value is too low, the product is too soggy. RF cooked comminuted samples were less hard (206.5 15.6 N) than WB cooked samples (302.2 35.4 N). The difference in springiness was not as significant as for the ground beef, however, it is still considerable. The RF samples exhibited significantly higher springiness (8.16 0.63 mm) as compared to the WB cooked samples (5.75 0.43 mm). Thus, RF cooked meat was more elastic than WB cooked. The gumminess of RF cooked (56.0 8.0 N) was lower than WB cooked (86.3 12.0 N) comminuted beef due to lower hardness values for RF cooked product. Overall, comminuted meat was successfully cooked with RF energy. The product was cooked through and exhibited adequate qualitative characteristics. Although the cooking time was approximately 7 or 8 min longer than the ground beef, the quality was much better. 3.5.3. Muscle beef product All TPA parameters were similar for RF and WB cooked beef muscle product. Thus RF and WB cooking provided similar textural quality in this product. This also indicates that RF is better in term of texture for whole muscle products (if uniformly cooked) than ground and comminuted beef products. Non-uniformity in cooking was observed due to development of air space between the product and container during cooking. It was due to the shrinkage of the product size. Different design of the cooking chamber is needed or the problem will be easily solved in continuous processing. 3.6. Temperature differential within the products Fig. 4 represents typical temperature history results observed with RF heating of all meat products in this study. The results from this study indicate that the product near the surface heats at a faster rate than in the centre. All trials demonstrated this pattern with ground
beef having a temperature differential of 12.2 2.3 C. The comminuted meat had a temperature differential of 19.7 9.3 C. The muscle had a temperature differential of 10.7 3.6 C. The muscle and ground beef cooked more uniformly than the comminuted meat. This may be due to non-uniform salt dispersion within the product. The portions of the meat with a higher salt concentration would have a higher dielectric constant. Piyasena, Dussault, Koutchma, Ramawamy, and Awuah (in press) illustrated the relationship between the dielectric properties of a product and the penetration depth. Since dielectric constant is inversely proportional to the penetration with the addition of salt, the penetration depth will decrease. Bengtsson and Risman (1971) reported a decrease in penetration depth with the addition of salt as did Bengtsson and Green (1970) when pasteurizing cured hams with RF.
4. Conclusions RF cooking of processed meat products resulted in decreased cooking time, lower juice losses, acceptable colour, WHC and texture. The ground beef resulted in the highest power efficiency but the texture was too chewy and elastic. An increased cooking time could improve the texture of the product but would reduce the efficiency. The temperature differential was acceptable for the ground and muscle meat. The comminuted and muscle meat products exhibited average energy efficiency with improved texture. The comminuted meat displayed a large cross-sectional temperature differential, possibly due to uneven salt distribution. The muscle’s colour was inferior. The well mixed comminuted and ground meat products appeared to be most promising for RF cooking. However, with specifically designed RF generators and applicator circuits, any one of these three products could be successfully processed and implemented in the industry.
Acknowledgements Funding for this project was provided by the University of Guelph, Natural Sciences and Engineering Research Council (NSERC) of Canada, and Agriculture and Agri-Food Canada. Authors appreciate technical assistance of C. Delifice in conducting this study. Contribution No. S127 from the Food Research Program. References
Fig. 4. Temperature–time relationship for comminuted beef cooked by RF.
Anon. (1996). Cooking with radio frequency. Food in Canada, 56(3), 10–33. Awuah, G. B., Ramaswamy, H. S., & Piyasena, P. (2002). Radio frequency (RF) heating of starch solution under continuous flow conditions: effect of system and product parameters on temperature
L. Laycock et al. / Meat Science 65 (2003) 959–965 change across the applicator tube. Journal of Food Process Engineering, 25, 201–223. Bengtsson, N. E., & Green, W. (1970). Radio frequency pasteurization of cured hams. Journal of Food Science, 35, 681–687. Bengtsson, N. E., & Risman, P. O. (1971). Dielectric properties of foods at 3 GHz as determined by cavity pertubation technique. Journal of Microwave Power, 6(2), 107–123. Houben, J., Schoenmakers, L., van Putten, E., van Roon, P., & Krol, B. (1991). Radio-frequency pasteurization of sausage emulsions as a continuous process. Journal of Microwave Power and Elecromagnetic Energy, 26(4), 202–205. McCormick, R. (1988). Dielectric heat seeks low moisture applications. Prepared Foods, 162(9), 139–140. Mittal, G. S., Nadulski, R., Barbut, S., & Negi, S. C. (1992). Textural profile analysis test conditions for meat products. Food Research International, 25, 411–417. Orsat, V., Bal, L., & Raghavan, G. S. V. (1999). Radio-frequency pasteurization of ham to enhance shelf life in vacuum packaging. ASAE/CSAE-SCGR Annual International Meeting, 18–22 July.
965
Piyasena, P., Dussault, C, Koutchma, T, Ramawamy, H. S., & Awuah, G. Radio frequency heating of foods: principles, application and related properties—a review. Critical Review in Food Science and Nutrition (in press). SAS. (1996). PC-SAS Users Guide, version 6. Cary, NC: SAS Institute. Strayfield. (1993). User manual: 1.5 kW radio frequency tube heater. Reading, UK: Strayfield. van Roon, P. S., Houben, J. H., Koolmees, P. A., Krol, B., & van Vliet, T. (1994). Mechanical and microstructural characteristics of meat doughs, either heated by a continuous process in a radio-frequency fields or conventionally in a water bath. Meat Science, 38, 103–116. Zhao, Y., Flugstad, B., Kolbe, E., Park, J. W., & Wells, J. H. (2000). Using capacitative (radio frequency) dielectric heating in food processing and preservation—A review. Journal of Food Process and Engineering, 23, 25–55. Zhang, M., Mittal, G. S., & Barbut, S. (1995). Effects of test conditions on the water holding capacity of meat by a centrifugal method. Lebensm.-Wiss. u.-Technol., 28, 50–55.
Radio frequency cooking of ground, comminuted and muscle meat products L. Laycocka, P. Piyasenab, G.S. Mittala,* a School of Engineering, University of Guelph, Guelph, Ontario, Canada N1G 2W1 Food Research Program, Agric & Agri-Food Canada, Guelph, Ontario, Canada N1G 2W1
b
Received 25 September 2002; received in revised form 12 November 2002; accepted 12 November 2002
Abstract The effect of radio frequency (RF) cooking, on the quality (colour, water holding capacity, texture), heating rate, and temperature history of three types of meat products (ground, comminuted and muscle), was investigated after heating to centre temperature of 72 C in a 1.5 kW RF heater operating at 27.12 MHz. The results were compared with that obtained from heating in a water bath. RF cooking of processed meat products resulted in a decreased cooking time (5.83, 13.5, and 13.25 min for ground beef, comminuted meat, and muscle, respectively compared to 151, 130, and 109 min in water bath), lower juice losses, acceptable colour, water holding capacity and texture. The results indicate that when using RF, ground beef has the highest power efficiency (60.17%) followed by comminuted meat (44.41%), and muscle (43.38%). However, the texture of ground beef was too chewy and elastic. The muscle’s colour was inferior. The comminuted and muscle meat products exhibited average energy efficiency with improved texture. The comminuted meat displayed a large cross-sectional temperature differential, possible due to uneven salt distribution. The well mixed comminuted and ground meat products appeared to be most promising for RF cooking. # 2003 Elsevier Ltd. All rights reserved. Keywords: Processed meat; Meat cooking; Meat heating; Meat quality; Meat colour; Meat texture; Texture profile analysis
1. Introduction The safety of proper cooking of meat and meat products is of increasing concern. Conventional food heating methods require that food be heated externally through conduction, convection or radiation. The main disadvantages of conventional methods are longer cooking time and non-uniform heating. The cooking time is lengthy for conventional methods as compared to dielectric heating such as radio frequency (RF) or microwave. With RF, due to volumetric heating, the product is heated internally, and not through the surfaces. The product forms a dielectric between the two electrodes acting as capacitor plates that are alternatively charged from positive to negative several millions times a second (27 MHz). The polar molecules in the product try to realign themselves with the polarity of the electrodes, causing a great deal of internal friction * Corresponding author. Tel. +1-519-824-4120x2431; fax: +1-519836-0227. E-mail address: [email protected] (G.S. Mittal). 0309-1740/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0309-1740(02)00311-X
to occur by which the product is heated. Also, RF cooking generally occurs uniformly across the sample and prevents the occurrence of degradation at the outside surfaces (Zhao, Flugstad, Kolbe, Park, & Wells, 2000). RF heating is currently practiced in the baking industry to remove excess moisture during post baking. Thawing of meats is also a common practice using RF technology (Anon., 1996; McCormick, 1988). Cooking and/or pasteurization of only a few selected meat products have been investigated. Houben, Schoenmakers, van Putten, van Roon, and Krol (1991) reported RF heating to be the most promising technique for the pasteurization of moving sausage emulsions in terms of penetration depth, energy efficiency and product quality. The pasteurization of cured hams was investigated by Bengtsson and Green (1970) and compared with conventional hot water treatments. RF treatment at 60 MHz reduced the processing time by one third, reduced juice losses and improved sensory quality of the meat. The only disadvantage was the increased bacterial count after three weeks of storage. Orsat, Bal, and Raghavan
960
L. Laycock et al. / Meat Science 65 (2003) 959–965
(1999) also used RF to pasteurize ham and found that the shelf life can be improved by using RF energy. Procter Strayfield at Horsham, PA developed a ‘‘Magnatube’’ pasteurization system capable of cooking scrambled eggs and skinless meatloaf (Zhao et al., 2000). Uniform temperature distribution and a high yield were achieved. The objective of this study was to evaluate the influence of RF cooking on the quality of each of the three types of meat products (ground, comminuted and muscle) after being heated to a centre temperature of 72 C.
2. Materials and methods 2.1. Equipment The experiment was performed at a constant frequency of 27.12 MHz with a heater of 1.5 kW (Strayfield Ltd, Reading, UK). The heater was comprised of two components: a RF generator and an applicator circuit (Awuah, Ramaswamy, & Piyasena, 2002; Strayfield, 1993). The applicator consisted of two electrodes, tuning capacitor plates, an inductance coil and an applicator cylinder. The cylinder was made of polytetrafluoroethylene (PFTE), which is RF inert. The cylinder (ID=0.0952 m, OD=0.121 m, height from inside=0.203 m, height from outside=0.215 m), with the product inside, was set between two electrodes. The electrodes wrapped around the applicator cylinder allowing a very small gap to remain between them. The electrodes were parallelograms and formed to wrap around the applicator cylinder such that every portion of the sample was subjected to dielectric heating. The electrodes had a diameter of 0.125 m and a height of 0.1496 m. The power from the RF generator was adjusted by changing the distance between the tuning plates of the tuning capacitor. If the plates are brought close together, the RF power applied to the product will increase and vise versa. This was accomplished manually by using a gear mechanism outside the RF oven (Strayfield, 1993). The active current passing through the product was maintained at constant level. When the product began to cook, the meat properties changed thus changing the active current. Active currents of 425 mA for ground beef, 320 mA for comminuted meat and 330 mA for whole muscle, were maintained by using the gear mechanism. The standing current, the current needed to maintain oscillations of the RF field when the applicator cylinder is empty, was 175 mA. The actual current is the difference between the two, resulting in 250, 145 and 155 mA. The power input to the RF heater is the result of the actual current multiplied by the high voltage. The high voltage for this unit was approximately 4.8 kV (Strayfield, 1993) resulting in a power input (Ps)
of 1200, 696 and 744 W, respectively, for ground beef, comminuted meat and whole muscle in this study. 2.2. Meat products Three types of meat products were prepared; ground beef, 2% salt comminuted beef and beef muscle. The ground beef used for both the ground and comminuted samples, was obtained from a local meat plant. The beef muscle product was obtained from Rowe Farm Meats, Guelph Ontario, Canada. The comminuted sample was prepared by evenly spreading 2% NaCl by mass around the sample then passing the product through a grinder (Hobart, Toronto, Canada; model #4532) with a 10 mm dia. openings twice. The muscle was cut into cylindrical pieces that would easily pack into the RF applicator cylinder. All raw samples had the same dimensions: a diameter of 0.0952 m and a height of 0.1496 m. The ground and comminuted meats were initially mixed thoroughly for 1 min with an electric mixer. In a 4 C refrigerator, the samples were layered into the custom made applicator cylinder, followed by a firm packing action after each layer to remove any air pocket that may have formed. The raw ground beef sample (1.15 0.0467 kg) contained 56% moisture, 27% protein, 16% fat and 1% ash. The raw comminuted sample (1.11 0.01 kg) contained 60% moisture, 24% protein, 13% fat, and 3% ash. The raw muscle sample (1.145 0.015 kg) contained 72% moisture, 22% protein, 5% fat and 1% ash. 2.3. Cooking and temperature measurement Two fiber-optic temperature probes (model # SFW-2, Luxtron Corporation, Santa Clara, CA) were calibrated using an oil bath. The sensors were connected to a signal converter and a display unit (755 multi-channel fluoroptic thermometer, Luxtron Corporation, Santa Clara, CA). One probe recorded the centre temperature (75 mm from the top surface) while the other measured the temperature 10–15 mm from the side (75 mm from the top surface). The temperature was manually recorded every 15 s until the centre temperature reached 72 C at which point the RF generator was ceased. The juice surrounding the product was removed and the product was weighed to determine the juice loss. The samples were then refrigerated at 4 C overnight. Three replications of each product were conducted. These products were also cooked in a water bath (VWR International, Sheldon Manufacturing, Cornelius, OR; model #1217). The dimensions of these samples were similar with a 9.5 cm diameter and 12 cm height. They were packed inside a plastic bag to prevent leakage and a sheet of aluminum foil to maintain their shape. The height of the water was about 12 cm inside the water bath, and the aluminum foil kept open at the
L. Laycock et al. / Meat Science 65 (2003) 959–965
top for the insertion of the temperature probe. One temperature probe (Omega, Laval, Canada; model HH23, microprocessor thermometer, type T thermocouple) was inserted in the product centre. Initially, the temperature of the water bath was 65 C and was gradually increased (0.2 C/min) to a final temperature of 85 C. The temperature was recorded every few minutes to obtain a temperature-time history. The product was removed from the water bath when the center temperature reached 72 C.
961
#5.0.1 was used to record the data. The results were then analyzed using a Microsoft Excel spreadsheet to compare the RF results to those measured using the water bath. 2.7. Data analysis All treatments were carried out three times. Duncan analysis for ranking means was conducted using the Statistical Analysis System (SAS, 1996).
2.4. Colour 3. Results and discussion The colour test measured the surface colour of the cooked meat with respect to lightness (L*), redness (a*) and yellowness (b*). For meat products, L* and a* are the most important parameters. The lighter and redder the product, the better the quality. A Minolta colorimeter (model CR-310, Minolta Co., Ramsey, NJ) and data processor were first calibrated with a known light source. Three colour replications of each cooked crosssection were then taken. 2.5. Water holding capacity (WHC) The centrifugal method of the water holding capacity (WHC) test measured the ability of the meat to absorb water. A product with a greater WHC is juicier and is generally preferred. Three replications of 5 g samples of each cooked product were placed inside centrifuge tubes while 8 ml of water was added to each tube. The tubes containing samples were incubated at 20 C for 1 h while the rotor was cooled in a refrigerator to 5 C. The centrifugation occurred at 7500 rpm for 7.5 min at 20 C (Zhang, Mittal, & Barbut, 1995). The WHC was then calculated using:
3.1. Cooking time The ground beef samples were cooked in 5.83 0.417 min using RF. The comminuted product was uniformly cooked in an average time of 13.5 2.0 min. The muscle was cooked in 13.25 1.75 min. The samples cooked in the water bath (WB) took much longer. The cooking times for the ground, comminuted and muscle samples were 151 1, 130 2 and 109 2 min, respectively. Cooking time in WB could be reduced by using faster rate of water temperature increase or higher initial water temperature. Figs. 1–3 compare the cooking times for ground, comminuted and muscle meat, respectively. The time taken to cook a 1.15 kg meat sample ranges from 5 to 15 min using RF energy compared to WB processing times of 107–130 min. Thus same amount of meat can be cooked approximately six times faster using RF. 3.2. Energy consumption and efficiency The power requirement of the process was calculated by determining the power required by the product over
WHC ¼ ðsolution added to the sample ðgÞ solution removed after centrifugation ðgÞÞ 100=ðmeat sample mass ðgÞÞ 2.6. Textural profile analysis (TPA) Texture profile analysis (TPA) measured the hardness, cohesiveness, springiness, gumminess and chewiness of a product. A universal testing machine (Instron, Burlington, Canada; model # 4204) was used with a 50 kN load cell. All samples were tested at room temperature having a diameter (D) of 25.5 mm and height (L) of 17 mm, resulting in a D/L ratio of 1.5. The samples were compressed down 12.75 mm or 75% (Mittal, Nadulsky, Barbut, & Negi, 1992) at a crosshead speed of 100 mm/ min. The test ran for the duration of two compression cycles. Labview data acquisition software, specifically custom-made interface software called Adlab version
Fig. 1. Temperature–time history of cooking ground beef (GB) using RF and a hot water bath (WB).
962
L. Laycock et al. / Meat Science 65 (2003) 959–965 Table 1 Mean colour parameters of both RF and water bath (WB) cooked meat products Product type
Cooking medium
Mean lightness (L*)
Mean redness (a*)
Ground beef Ground beef
RF WB
44.261.10a 43.191.45a
8.22 0.93a 8.85 0.30a
Comminuted Comminuted
RF WB
42.791.06b 41.420.44b
7.33 1.13b 7.93 1.22b
Muscle Muscle
RF WB
49.831.86c 52.500.71c
7.35 0.62b 7.96 0.14b
Means with the same letter in a column are not different at 95% level.
Fig. 2. Temperature–time history of cooking comminuted (COM) beef using RF and a hot water bath (WB).
Fig. 3. Temperature–time history of cooking beef muscle using RF and a hot water bath (WB).
the entire cooking time. The power required can be calculated using: Pw ¼ mcp T=t where Pw is the power required, W, cp is the specific heat, J/(kg C), m is the mass, kg, T is the temperature change, C, and t is the cooking time, s. The overall power efficiency can be calculated using (Houben et al., 1991): Pefficiency ¼
Pw 100 Ps
where Pw is the power required by the product, and Ps is the power supplied to the RF generator. The results indicate that when using RF, ground beef has the highest power efficiency of 60.17 2.26%. Comminuted meat has the next highest power efficiency
of 44.41 6.27%. Muscle product resulted in a power efficiency of 43.38 5.17%. Bengstsson and Green (1970) pasteurized ham with RF energy and reported the highest power efficiency of 25%. It was noted that the requirement for an industrial application would have to be 50–60% efficiency. Houben et al. (1991) performed pasteurization of moving sausage emulsions and reported average power efficiencies of 25–30%, which could be improved by specifically designing the generator-heating unit combination. Zhao et al. (2000) suggested that the energy losses within an RF system include matching networks, couplers and the capacitive electrode system. Therefore, overall energy efficiencies would probably not exceed 50–60% with the present technologies available. The results of this study indicate that RF energy is much more energy efficient as compared to WB processing. The energy efficiency of ground beef using RF was 48.13 1.81% as compared to WB heating results of 2.53 0.06%. The comminuted meat cooked with RF had an energy efficiency of 20.61 2.91% as compared to the WB cooked samples of 2.53 0.06%. The muscle RF cooked samples had an energy efficiency of 21.52 2.57% as compared to the WB samples of 2.61 0.07%. The low efficiency in case of WB cooking was due to the cooking of small amount of meat product in a large amount of water and large losses of heat to environment. Hence this is not the true efficiency achieved at industrial scale. Nevertheless, it is much more efficient to use RF as opposed to WB processing to cook all three meat products. Altering the design of the generator or creating a continuous flow tube could improve the energy efficiencies of both RF cooked comminuted and muscle meats. 3.3. Colour The results from the colour test are summarized in Table 1. In terms of L* and a*, the RF cooked meat products samples were similar to those WB cooked. The differences were so small that these were also not visible to the naked eye.
963
L. Laycock et al. / Meat Science 65 (2003) 959–965
3.4. Water holding capacity (WHC) Table 2 summarizes the results of water holding capacity using the centrifugal method. RF comminuted meat did not absorb as much water as compared to WB cooking. This may cause the meat to be little drier to the palate but may be acceptable. The RF muscle samples have a much better WHC as compared to the WB cooked samples. In terms of WHC, the results do not suggest that RF energy cause the meat to become less absorbent and in turn less juicy. Juice loss during cooking is another method of determining the juiciness of a product. Table 3 summarizes
the juice loss percentage of both RF and WB cooked products. The RF cooked samples seemed to hang on to their moisture during cooking. This would have a lot to do with cooking time, the longer the cooking time, the drier a product becomes. These results suggest that RF heating reduces moisture loss and the result is a juicier product. This finding has also been reported earlier (Anon., 1996; Bengtsson & Green, 1970). Moisture content was similar for both cooking methods (WB and RF) suggesting that the WB cooked products have a tendency to lose more fat than water during cooking, keeping the water content at the same level. 3.5. Texture profile analysis
Table 2 Mean water holding capacity (WHC) values of both RF and water bath (WB) cooked meat products Product type
Cooking medium
Water holding capacity
Ground beef Ground beef
RF WB
6.630.09b 6.610.05c
Comminuted Comminuted
RF WB
6.630.02b 6.670.06a
Muscle Muscle
RF WB
6.610.08c 6.550.06d
Means with the same letter in a column are not different at 95% level. Table 3 Mean juice loss percentages of both RF and water bath (WB) cooked meat products Product type
Cooking medium
Juice loss (%)
Ground beef Ground beef
RF WB
25.5 3.7bc 26.0 0.1b
Comminuted Comminuted
RF WB
20.9 1.2d 23.4 0.6c
Muscle Muscle
RF WB
28.7 6.9ab 31.2 0.3a
Means with the same letter in a column are not different at 95% level.
The TPA parameters of ground beef, comminuted meat, and muscle cooked with RF and WB are shown in Table 4. 3.5.1. Ground beef The mean hardnesses (1 and 2), cohesiveness and gumminess were similar for RF and WB cooked ground beef. The springiness represents the elastic properties of the meat. The higher the springiness, the more elastic or rubbery characteristics the product will exhibit. The ground beef cooked with RF had significantly higher springiness (9.85 0.56 mm) than WB cooked (4.91 0.55 mm). Thus, the RF samples were more elastic, and may contribute to the product chewiness. A large difference was observed with respect to chewiness. The chewiness value of the RF cooked ground beef (926.8 171.9 mJ) was significantly higher compared to WB cooked (430.1 103.7 mJ). The product cooked with RF was much too chewy. Thus, RF cooked ground beef was significantly more elastic and chewy. Van Roon, Houben, Koolmees, Krol, and van Vliet (1994) also reported that meat dough heated by RF was much firmer than those WB heated. The potential for RF cooking of ground beef looks promising, however, slightly slower heating is required and better process conditions are needed for the integration of such a process in industrial practices.
Table 4 Mean and standard deviation of TPA parameters of RF and WB cooked ground, comminuted and muscle beef products Meat product
Cooking medium
Hardness1 (N)
Hardness2 (N)
Cohesiveness
Springiness (mm)
Ground Ground
RF WB
313.337.2b 32934.3b
225.831.0b 238.428.5b
0.290.02b 0.270.02b
9.850.56a 4.910.55d
926.8171.9b 436.1103.7c
93.814.7 87.813.3b
Comminuted Comminuted
RF WB
206.515.6c 302.235.4b
152.514.9c 233.933.0b
0.270.02b 0.280.01b
8.160.63b 5.760.43d
459.991.0c 497.679.8c
568.0c 86.312.0b
Muscle Muscle
RF WB
709.2143.8a 698.2126.8a
588.1122.8a 592.6118.3a
0.430.04a 0.390.06a
6.60.68cd 7.260.28c
2059.1531.3a 2001523.9a
311 70.0a 275.770.1a
Means with the same letter in a column are not different at 95% level.
Chewiness (mJ)
Gumminess (N)
964
L. Laycock et al. / Meat Science 65 (2003) 959–965
3.5.2. Comminuted beef Average values of cohesiveness and chewiness were similar for RF and WB cooked comminuted beef. If the hardness value is too high, the product is too tough. If the hardness value is too low, the product is too soggy. RF cooked comminuted samples were less hard (206.5 15.6 N) than WB cooked samples (302.2 35.4 N). The difference in springiness was not as significant as for the ground beef, however, it is still considerable. The RF samples exhibited significantly higher springiness (8.16 0.63 mm) as compared to the WB cooked samples (5.75 0.43 mm). Thus, RF cooked meat was more elastic than WB cooked. The gumminess of RF cooked (56.0 8.0 N) was lower than WB cooked (86.3 12.0 N) comminuted beef due to lower hardness values for RF cooked product. Overall, comminuted meat was successfully cooked with RF energy. The product was cooked through and exhibited adequate qualitative characteristics. Although the cooking time was approximately 7 or 8 min longer than the ground beef, the quality was much better. 3.5.3. Muscle beef product All TPA parameters were similar for RF and WB cooked beef muscle product. Thus RF and WB cooking provided similar textural quality in this product. This also indicates that RF is better in term of texture for whole muscle products (if uniformly cooked) than ground and comminuted beef products. Non-uniformity in cooking was observed due to development of air space between the product and container during cooking. It was due to the shrinkage of the product size. Different design of the cooking chamber is needed or the problem will be easily solved in continuous processing. 3.6. Temperature differential within the products Fig. 4 represents typical temperature history results observed with RF heating of all meat products in this study. The results from this study indicate that the product near the surface heats at a faster rate than in the centre. All trials demonstrated this pattern with ground
beef having a temperature differential of 12.2 2.3 C. The comminuted meat had a temperature differential of 19.7 9.3 C. The muscle had a temperature differential of 10.7 3.6 C. The muscle and ground beef cooked more uniformly than the comminuted meat. This may be due to non-uniform salt dispersion within the product. The portions of the meat with a higher salt concentration would have a higher dielectric constant. Piyasena, Dussault, Koutchma, Ramawamy, and Awuah (in press) illustrated the relationship between the dielectric properties of a product and the penetration depth. Since dielectric constant is inversely proportional to the penetration with the addition of salt, the penetration depth will decrease. Bengtsson and Risman (1971) reported a decrease in penetration depth with the addition of salt as did Bengtsson and Green (1970) when pasteurizing cured hams with RF.
4. Conclusions RF cooking of processed meat products resulted in decreased cooking time, lower juice losses, acceptable colour, WHC and texture. The ground beef resulted in the highest power efficiency but the texture was too chewy and elastic. An increased cooking time could improve the texture of the product but would reduce the efficiency. The temperature differential was acceptable for the ground and muscle meat. The comminuted and muscle meat products exhibited average energy efficiency with improved texture. The comminuted meat displayed a large cross-sectional temperature differential, possibly due to uneven salt distribution. The muscle’s colour was inferior. The well mixed comminuted and ground meat products appeared to be most promising for RF cooking. However, with specifically designed RF generators and applicator circuits, any one of these three products could be successfully processed and implemented in the industry.
Acknowledgements Funding for this project was provided by the University of Guelph, Natural Sciences and Engineering Research Council (NSERC) of Canada, and Agriculture and Agri-Food Canada. Authors appreciate technical assistance of C. Delifice in conducting this study. Contribution No. S127 from the Food Research Program. References
Fig. 4. Temperature–time relationship for comminuted beef cooked by RF.
Anon. (1996). Cooking with radio frequency. Food in Canada, 56(3), 10–33. Awuah, G. B., Ramaswamy, H. S., & Piyasena, P. (2002). Radio frequency (RF) heating of starch solution under continuous flow conditions: effect of system and product parameters on temperature
L. Laycock et al. / Meat Science 65 (2003) 959–965 change across the applicator tube. Journal of Food Process Engineering, 25, 201–223. Bengtsson, N. E., & Green, W. (1970). Radio frequency pasteurization of cured hams. Journal of Food Science, 35, 681–687. Bengtsson, N. E., & Risman, P. O. (1971). Dielectric properties of foods at 3 GHz as determined by cavity pertubation technique. Journal of Microwave Power, 6(2), 107–123. Houben, J., Schoenmakers, L., van Putten, E., van Roon, P., & Krol, B. (1991). Radio-frequency pasteurization of sausage emulsions as a continuous process. Journal of Microwave Power and Elecromagnetic Energy, 26(4), 202–205. McCormick, R. (1988). Dielectric heat seeks low moisture applications. Prepared Foods, 162(9), 139–140. Mittal, G. S., Nadulski, R., Barbut, S., & Negi, S. C. (1992). Textural profile analysis test conditions for meat products. Food Research International, 25, 411–417. Orsat, V., Bal, L., & Raghavan, G. S. V. (1999). Radio-frequency pasteurization of ham to enhance shelf life in vacuum packaging. ASAE/CSAE-SCGR Annual International Meeting, 18–22 July.
965
Piyasena, P., Dussault, C, Koutchma, T, Ramawamy, H. S., & Awuah, G. Radio frequency heating of foods: principles, application and related properties—a review. Critical Review in Food Science and Nutrition (in press). SAS. (1996). PC-SAS Users Guide, version 6. Cary, NC: SAS Institute. Strayfield. (1993). User manual: 1.5 kW radio frequency tube heater. Reading, UK: Strayfield. van Roon, P. S., Houben, J. H., Koolmees, P. A., Krol, B., & van Vliet, T. (1994). Mechanical and microstructural characteristics of meat doughs, either heated by a continuous process in a radio-frequency fields or conventionally in a water bath. Meat Science, 38, 103–116. Zhao, Y., Flugstad, B., Kolbe, E., Park, J. W., & Wells, J. H. (2000). Using capacitative (radio frequency) dielectric heating in food processing and preservation—A review. Journal of Food Process and Engineering, 23, 25–55. Zhang, M., Mittal, G. S., & Barbut, S. (1995). Effects of test conditions on the water holding capacity of meat by a centrifugal method. Lebensm.-Wiss. u.-Technol., 28, 50–55.