Microwave Drying to Determine the Solids Content of Milk and Cottage and Cheddar Cheese

Transcription

1 7 Journal of Food Protection, Vol. 7, No., Pages 7-78 (April 98) Copyright', International Association of Milk, Food, and Environmental Sanitarians Microwave Drying to Determine the Solids Content of Milk and Cottage and Cheddar Cheese DAVID M. BARBANO* and MARY E. DELLA VALLE Department of Food Science, Cornell University, Ithaca, New York 8 (Received for publication September 9, 98) ABSTRACT Microwave ovens have been used in quality control laboratories for rapid determination of moisture content of dairy products. Factors that influence sample drying rate and final test result for microwave drying are quite different than factors that influence conventional drying methods. Differences in dielectric properties of samples will have a significant influence on microwave drying, but may not influence conventional drying. Sample handling methods that reduce variability in test results for milk, cottage cheese, and Cheddar cheese are discussed. Preliminary investigations indicated that there are differences in results from one microwave oven to the next, even when the same sample material was dried under the same analysis conditions. This variations may be due to a combination of differences in sample positioning, differences in wave patterns in individual oven cavities, and differences in magnetron power output with age and heavy usage. Good correlations of microwave test results with those from the standard method were obtained for milk and cottage and Cheddar cheese. Statistically significant differences between microwave results and standard method results indicated that for best agreement between laboratories it would be necessary to calibrate individual microwave ovens against a standard reference method for each type of product. Determination of the moisture content of food and dairy products is one of the most common tests done in quality control laboratories. Product moisture determinations are routinely done for reasons of regulatory compliance and maintenance of product quality and product yield. In other instances, such as the determination of solids content of milk, results may be used as a basis of payment to the farmer (). Even though water is one of the major constituents of most food products, it is a substance that is very difficult to accurately quantitate. Karmas () reviewed various methods of moisture determination in foods. Gravimetric methods are the most common methods used for moisture determination in food products. Forced air or vacuum ovens are commonly used to drive moisture out of a sample. In many applications, there is a need for rapid determination of moisture for on-line product quality control. Microwave ovens have the ability to heat samples very quickly and drive off their moisture in to min. Previous studies have demonstrated the application of microwave drying to milk (,,0) and cheese (7,8). Pieper (7) did a collaborative study on microwave moisture analysis of Cheddar-style cheeses. The Methods of the Association of Official Analytical Chemists () define specific conditions of analysis and specify a model of microwave oven that is no longer commercially available. The objective of our study was to investigate some of the common problems encountered in microwave moisture determination and evaluate the repeatability, reproducibility, and accuracy of microwave equipment currently being used by the dairy industry. MATERIALS AND METHODS Sample handling Two sets of milk and cottage and Cheddar cheese samples were delivered to each cooperating laboratory. Out-of-state laboratories received their samples by overnight air freight delivery. oratories in New York State had samples delivered directly on the same day as the out-of-state laboratories. All samples were packed in ice during shipping. One set, samples of each type of product, was sent to each laboratory, and then a second set of different samples of each product was sent to each laboratory about weeks later. All laboratories were contacted before the collaborative study and encouraged to obtain milk and cottage and Cheddar cheese before the collaborative study to become familiar with handling these types of samples if they were not routinely analyzing these products. Preliminary practice samples allowed cooperating laboratories to evaluate what combination of power, time, and sample size gave reproducible results on their particular microwave oven. All of the cooperating laboratories were routinely using a microwave oven to analyze at least one of the three types of dairy products. All laboratories were instructed to do the samples in order to and to do each sample in duplicate before going on to the next sample. Results were reported to two decimal places. Each laboratory reported the type of oven used and the conditions of analysis for each type of product. Milk samples Raw whole milk used in this study was obtained from 0 individual cows at the Cornell University dairy farm. Total solids content of all 0 samples was determined in duplicate by AOAC method.0 (), using a steam bath for predrying and a forced air oven (Fisher Model 0F) set at C exit air temperature. Temperature uniformity in this oven was checked and found to be within C of the exit air temperature at all sample locations within the oven. Samples were weighed to four decimal places on an analytical balance that was calibrated with Class S standard calibration weights. JOURNAL OF FOOD PROTECTION, VOL. 7, APRIL 98

2 DETERMINING MOISTURE IN MILK AND CHEESE 7 Twelve of the 0 samples were selected to give a good range of total solids content for the study. Samples were mixed and split for distribution to collaborators. Collaborators were given the following instructions for sample handling and analysis. Refrigerate all samples upon arrival. All milk samples are raw non-homogenized milk. Proper sample mixing is extremely important. Pour each sample back and forth between the original sample container and another clean and dry container at least times before pipetting ml of sample onto two tared glass fiber pads. If the sample has a cream line on the inside of the original container, it may be necessary to warm the sample to room temperature to obtain complete mixing. It is best to pipette the sample on to the center of the pad and let it absorb the milk instead of spotting it around in different size drops in different locations. The duration of the microwave exposure should be at least min, but can be longer if the analyst finds that longer times give more reproducible results. The appropriate microwave power setting is to be determined by the analyst. Select a power setting that gives minimal sample browning and reproducible results. Use this same power setting and time for all milk samples. Collage and Cheddar cheese samples Cottage cheese (% milkfat) was obtained from one commercial cheese plant. Different samples were obtained from different individual vats of cheese over a -d period to obtain a good range of sample moistures. All samples were blended to a smooth creamy consistency in a heavy duty commercial blender and split into portions for each laboratory. This was done to obtain a homogeneous sample that could be split into identical sub samples for distribution. Cheddar cheese samples were obtained from one commercial cheese plant. Samples from different vats of cheese were selected to give a wide range of moisture levels. All cheese samples were ground in a blender (- to -mm particle size), thoroughly mixed, and split into sample containers for distribution. The sample containers were completely filled to eliminate any open head space above the cheese. This minimized loss of moisture from the cheese at the top of the container. oratories were given the following instructions about cheese sample handling and analysis. Samples should be kept refrigerated. Select a sample size within the range of to g and use that sample weight for all analyses. Duration of microwave exposure should be at least min but can be longer if the analyst finds that a longer time gives more reproducible results. Selection of an appropriate microwave power setting is at the discretion of the analyst. Select a combination of power, time, and sample size that gives reproducible results. Do cheese samples between two glass fiber pads to minimize any spattering that might occur. It is best to place all of the sample at the center of the pad in a - to -cm circle. All of the sample should be in one continuous mass of - to -cm diameter and not in discrete spots or pieces distributed over the surface of the glass fiber pad. This will result in a more uniform final sample temperature and improve repeatability. Cottage cheese was analyzed by the AOAC vacuum oven method ()., using a Fisher Isotemp vacuum oven. Cheddar cheese was analyzed by forced air oven drying of a -g sample for hat C. Statistical methods Repeatability and reproducibility were calculated by the methods of Youden and Steiner (). Means, standard deviations, and regression analyses were done using Minitab. Analysis of variance and Duncan's Multiple Range Test were calculated using SAS. Duncan's Multiple Range Test was only done if a statistically significant F value was obtained by the analysis of variance (9). RESULTS AND DISCUSSION Preliminary studies Microwave heating and drying is fundamentally different from the application of heat to samples in a convection oven. In a convection oven, the surface of the sample is heated by convection of heated oven air. The hot walls or heating elements radiate infrared waves which are absorbed at the surface of the sample. The interior of the sample is then heated by conduction from the heated sample surface. In microwave heating the microwaves penetrate and are absorbed more uniformly through the sample. Microwave energy is converted into molecular motion and thermal energy within the sample. This causes very rapid and uniform heating. Factors that influence the microwave absorptivity of a particular type of sample material will influence microwave moisture determinations. Preliminary work with cottage cheese gave a very good example of this problem. Samples of cottage cheese curd were taken from a cheese vat before cream dressing was applied to the curd. After cream dressing was added to the batch of cheese curd, more samples were taken. The uncreamed cottage cheese curd dries very slowly in a microwave oven. In contrast, creamed cottage cheese from the same vat dries much more rapidly, under the same conditions of analysis in a microwave oven. The critical difference is not the fact that the uncreamed curd was 0.% milk fat while creamed cottage cheese has % milkfat. The most important difference between the two types of samples is the salt content. Cream dressing that is added to the curd contains added salt. Added salt influences the dielectric properties of the aqueous phase of the cottage cheese. Salt increases the microwave absorptivity of the cottage cheese, and thus the creamed curd cottage cheese will dry faster and more completely in a microwave oven than uncreamed cottage cheese curd. The physical presentation of the sample and the sample size are also important. If for example, g of Cheddar cheese are presented to the oven in two very dfferent ways, the results may be completely different. If one portion of Cheddar cheese is compacted into a tight ball and squeezed between glass fiber pads versus spreading the same weight of cheese uniformily over the whole surface of the fiber pad, the results will be very different. The portion of the sample that is tightly compacted will not be able to dissipate heat as quickly as a sample that is spread out. The sample that is compacted will reach a higher internal temperature than the sample that was not compacted even though they were both exposed to the same amount of microwave radiation. In this example, you might find that the compacted sample would get so hot it would burn while the other sample would not completely dry under the same analysis conditions. With many types of samples it is difficult to spread the sample uniformly on the glass fiber pads. Our instructions for cheese had approximately the same weight of cheese spread over the same small areas of the fiber pad in each laboratory. It is best if all of the sample is in contact with itself so that a more uniform temperature is obtained in all particles of the sample. However, excessive compaction of the sample is not recommended because it will lead to sample burning. Another aspect of our preliminary work included a splitsample collaborative study with various dairy products. In this study, we had three different models of microwave ovens from different companies at the university laboratory. Extensive work was done to determine the sample size, power, and time that gave results that agreed with the AOAC method for each particular dairy product on each type of JOURNAL OF FOOD PROTECTION, VOL. 7, APRIL 98

3 7 BARBANO AND DELLA VALLE microwave oven. Subsequently, samples were sent out to cooperating laboratories. Exact analysis conditions were specified for each type of microwave oven. The results were very bad. In one laboratory, samples tested under the specified conditions would burn, whereas in another laboratory with the same model of microwave oven the samples did not dry completely. It appeared that the same numeric values for power intensity for given models of microwave oven gave significantly different actual power exposures to the samples. The differences may have been due to sample position effects () or differences in wave patterns of the individual microwave ovens (/). Some brands of microwave ovens had much more oven-to-oven variability than others. Based on our preliminary work, we conducted a collaborative study on microwave moisture determination for milk and cottage and Cheddar cheese as described in the Material and Methods. Instead of specifying time and power, we instructed each laboratory to select the operating conditions (for their particular oven) that would give good repeatability and reasonable results for the types of products to be studied. Collaborative study Milk analysis. Actual conditions of analysis used in each laboratory for each type of product are specified in Table. Results of the split-sample study for milk are shown in Tables,, and 8. The high/low total solids range of milk sample set was.7 to. and sample set was.0 to.. Results in Table indicate that the within-laboratory repeatability and between-laboratory reproducibility (77) were similar for both sets of milk samples. Within-laboratory repeatability was better than between-laboratory reproducibility (Table ). A very good correlation of analysis results obtained by microwave and the AOAC methods was observed (Table TABLE. Microwave oven analysis conditions used in each laboratory. Product Sample size (g) Percent power Time (min) Oven type" Milk Cottage cheese Cheddar cheese a Mention of trade names does not imply endorsement by Cornell University. TABLE. Average percent solids content of milk determined by microwave analysis and standard method. Sample set Sample set XM" Xs b XM-XS XM Xs a X M - mean microwave result. b X s - mean standard method result X M"Xs Microwave analysis Sample set Sample set Repeatability 0"o.0.0 Reproducibility.. JOURNAL OF FOOD PROTECTION, VOL. 7, APRIL 98

4 DETERMINING MOISTURE IN MILK AND CHEESE 7 ). However, the comparison of the combined results (Table 8) on all milk samples by Duncan's Multiple Range Test (9) indicates that all laboratories except number, obtained results that were significantly different (a =.0) than the standard method. Given the good linear relationship between microwave results and the standard method (Table ), it can be concluded [in agreement with previous results of Green ()] that microwave oven results can be adjusted by regression analysis to agree with the official method for milk. It is also noteworthy from the statistical analysis in Table 8, that there were significant differences between microwave results from different laboratories which used the same microwave oven operating conditions. This agreed with results from our preliminary work. Therefore, to obtain better between-laboratory reproducibility with this method, it would be necessary to calibrate the microwave oven by correlation with results from the standard method on split samples. Cheese analysis. Actual analysis conditions used in each laboratory for microwave total solids determinations on cottage and Cheddar cheese are given in Table. Average percent solids content for cottage cheese sets and are shown in Table. The high/low solids range for individual samples in set was 0.8 to 9.8 and set was 0. to.. Again, within-laboratory repeatability was better than between-laboratory reproducibility (Table ). Correlation between microwave results and the standard method were good (Table ), but not as good as those observed for milk (Table ). Variations in batch-to-batch formulation of different lots of cottage cheese would be expected to cause more variation in test results than those observed for milk. The comparison of the combined results on cottage cheese set and by Duncan's Multiple Range Test (9) indicated that the mean results by the microwave method were all significantly different from the standard method (Table 8). Only two of the six laboratories ( and ) had microwave results that were not significantly different from each other. Again, the results of Duncan's Multiple Range Test, plus the fact that the withinlaboratory repeatability was better than between-laboratory reproducibility, indicate that better between-laboratory agreement could be obtained by calibration of individual microwave ovens with samples of known solids content. Results of the Cheddar cheese analysis are shown in Table and 7. The repeatability and reproducibility were not as good as those observed for milk and cottage cheese. The high/low solids range of individual samples, as determined by the standard method, was.9 to.7 for set and. to 8. for set. Comparisons of microwave and standard method results of sample sets and by the Duncan's Multiple Range Test (9) indicated that the average results from all laboratories were significantly different from the standard method (Table 8). Calibration of the microwave ovens would be of value in reducing between-laboratory variability in results on Cheddar cheese analysis by microwave ovens. The time and labor involved to do extensive calibration of microwave ovens, with results from the standard method, has to be balanced against the accuracy needed in the final results. For routine quality control purposes, the differences such as those observed between Cheddar cheese results (Table 8) for laboratories,, and the standard method may TABLE. Regression analysis" of results of microwave and standard method determination of total solids content of milk. Sample " " " " " " " " " " "Simple linear regression of the form Y = A + BX; S B is the standard deviation of Y with Tespect to X; Y = microwave and X = standard method. "intercept not significantly different from zero. JOURNAL OF FOOD PROTECTION, VOL. 7, APRIL 98

5 7 BARBANO AND DELLA VALLE TABLE. Average percent solids content of cottage cheese determined by microwave analysis and standard method. Sample set Sample set XM-XS XM-XS a X M -mean microwave result. b X s -mean standard method result Microwave analysis Sample set Sample set Repeatability Co.0.0 Reproducibility.. TABLE. Regression analysis " of results of microwave and standard method Sample B.9 determination of total solids content of cottage cheese. A S B r.7 b b b b l.ll b b b b b b a Simple linear regression of the form Y = A + BX; S B is the standard deviation of Y with respect to X; Y = microwave and X = standard method. b Intercept not significantly different from zero. TABLE. Average percent solids content of Cheddar cheese determined by microwave analysis and standard method. Sample set Sample set XM-XS XM-XS a X M -mean microwave result. b X s -mean standard method result Microwave analysis Sample set Sample set Repeatability..09 Reproducibility..8 JOURNAL OF FOOD PROTECTION, VOL. 7, APRIL 98

6 DETERMINING MOISTURE IN MILK AND CHEESE 77 all be-statistically different from each other, but from a practical point of view may be close enough for quality control work. On the other hand, for products such as milk where the product price will be determined by the test result, more accurate calibration of a microwave oven would be a necessity. In products such as Cheddar and cottage cheese where formulations may differ from plant to plant, the behavior of one company's product may be very different from another company's product under the same analytical conditions in the same microwave oven. This is the reason that for this collaborative study, Cheddar and cottage cheeses were obtained from only one company. Thus company to company differences in product formulation would not be a confounding variable in our evaluation of microwave oven performance. It is not uncommon to see a range of salt content in commer- T ABLE 7. Regression analysis" of results of microwave and standard method determination of total solids content of Cheddar Sample SB.9.77".0 cheese " " " " " " " " " " ".0.99 a Simple linear regression of the form Y = A + BX; S B is the standard deviation of Y with respect to X; Y = microwave and X = standard method. "intercept not significantly different from zero. TABLE 8. Comparison variance. 'Means within Range Test. i Standard method of combined means of data from samples sets and for milk and cottage and Cheddar cheese and analysis of.". d.9 a. d.7 c. d Percent solids Microwave Milk Cottage Cheddar 9. d.9 a 9. e. e 9.".8" 9.8 a. d 9.8 f. f 9. e. f." product type followed by different superscripts are significantly different at p<.00 as determined by Duncan's Multiple 9.0 c.7 c Analysis of Variance Milk Cottage Cheddar Source DF F Value P R >F a F Value P R >F F Value P R >F X Sample a P R >F - probability of greater F value. JOURNAL OF FOOD PROTECTION, VOL. 7, APRIL 98

7 78 BARBANO AND DELLA VALLE rial Cheddar cheese samples from. to.% or in Mozzarella cheese from.0to.7%. Reduction in the variability in microwave results with respect to the standard method is limited by differences in sample composition that influence the dielectric properties of the aqueous phase of the sample. Like most analytical methods, microwave drying works best with samples that are physically and chemically homogeneous. It must be remembered that differences in chemical composition of samples that have little or no impact on tests results by traditional drying methods may have a significant impact on microwave drying. ACKNOWLEDGMENTS The authors thank Joni Ruller, Maureen Chapman, and Margaret Jacobs for their technical assistance in this study. The assistance and cooperation of the technical staff in the collaborating laboratories ( Corporation, Cuba Cheese, Dairy Chemistry oratory - California Department of Food and Agriculture, Friendship Dairies, and Kraft Inc.) was greatly appreciated. Financial support for this study by the American Public Health Association is gratefully acknowledged. REFERENCES. Cumutte, B Principles of microwave radiation. J. Food Prot. :8-.. Green, W. C, and K. K. Park Comparison of AOAC, microwave and vacuum oven methods for determining total solids in milk. J. Food Prot. : Hamada, H., M. Yoshino, K. Shiga, Y. Takahashi, A. Nagai, andn. Suga Rapid determination of milk solids by microwave heater andelectrobalance. J. Dairy Sci. 0:-7.. Hayward, L. H., and D. H. Kropf Sample position effects on moisture analyses by a microwave oven method. J. Food Prot. :- 7.. Horowitz, W. (ed.) Official methods of analysis of the Association of Official Analytical Chemists, th ed. Association of Analytical Chemists, Washington, DC.. Karmas, E Techniques for measurement of moisture content of foods. FoodTechnol. (): Pieper, H., J. A. Stuart, and W. R. Renwick Microwave technique for rapid determination of moisture in cheese. J. Assoc. Off. Anal. Chem. 0: Shanley,R. M., and G. W. Jameson. 98. A study of the rapid determination of moisture in cheese by microwave heating. Austral. J. DairyTechnol. : Snedecor, G. W., and W. G. Cochran Statistical methods, 7th ed. Iowa State University Press, Ames, Iowa. 0. Yoshino, M., H. Hamada, and Y. Takahashi Rapid determinations of total milk solids with a modified automated microwave oven andelectrobalance. J. Dairy Sci. :-7.. Youden, W. J., and E. H. Steiner. 97. Statistical manual of the Association of Official Analytical Chemists. Association of the Official Analytical Chemists, Washington, DC. JOURNAL OF FOOD PROTECTION, VOL. 7, APRIL 98