Physicochemical and Microbiological Properties of Selected Rice ...

Physicochemical and Microbiological Properties of Selected Rice Flour-Based Batters for Fried Chicken Drumsticks A. Mukprasirt,* T. J. Herald,*,1 D. L. Boyle,† and E. A. E. Boyle‡ *Food Science Program, Kansas State University, Manhattan, Kansas 66506; †Division of Biology, Kansas State University, Manhattan, Kansas 66506; and ‡Animal Sciences and Industry, Kansas State University, Manhattan, Kansas 66506 than WFBB. The TBA values of RFBB and WFBB increased (P < 0.05) with increased frozen storage time at −40 C for 90 d. The RFBB with MC exhibited the lowest TBA values, whereas WFBB had the highest values. Microstructural analysis revealed that freezing caused structural deterioration of all batters, but the RFBB with MC exhibited less freezing tolerance than other samples. The total plate counts of immediately fried or frozen fried chicken stored for 90 d were less than 1 log cfu/g sample. The RFBB with 5% oxidized corn starch and MC can replace WFBB on fried drumsticks. Additionally, RFBB results in a healthier product due to lower fat absorption.

ABSTRACT Rice flour-based batter (RFBB) formulations for chicken drumstick coating were developed as an alternative for traditional wheat flour-based batter (WFBB). Physicochemical properties and storage stability of selected RFBB were evaluated and compared to WFBB. Batter pickup of RFBB formulated in combination with oxidized corn starch and methylcellulose (MC) was not significantly different from that of WFBB. In contrast, batters with only rice and corn flour (60:40% flour weight) exhibited significantly higher pickup. Rice flour batter with 15% oxidized corn starch had the lowest batter pickup. All RFBB exhibited (P < 0.05) lower oil absorption

(Key words: chicken drumstick, rice flour, batter, coating, adhesion batter) 2001 Poultry Science 80:988–996

absorption of battered foods through several chemical and physical changes that occur, including starch gelatinization, protein denaturation, water vaporization, and crust formation (Saguy and Pintus, 1995). Oil absorption during deep-fat frying is a function of oil quality (AbdelAal and Karara, 1986; Blumenthal, 1991), composition (Makinson et al., 1987), and initial moisture content, product shape and porosity, and frying time and temperature (Pravisani and Calvelo, 1986; Gamble et al., 1987a,b; DuPont et al., 1992; Pinthus and Saguy, 1994). Rancidity development during storage is a major concern in frozen, fried chicken products. Factors influencing the lipid oxidation rate include fatty acid composition, oxygen concentration, temperature, surface area of lipids exposed to air, moisture content, and pro-oxidants (Nawar, 1996). Shih and Daigle (1999) reported that oil uptake of rice flour fried batter was less than that of wheat flour-based batter (WFBB). Mukprasirt et al. (2000a) developed a rice flour-based batter (RFBB) for fried chicken drumsticks

INTRODUCTION The consumption of battered or breaded chicken has increased significantly, and annual sales of fried chicken in United States during 1995 were estimated to be more than six billion dollars (Mohan Rao and Delaney, 1995). Commercially, battered or breaded foods are fully or partially cooked by deep-fat frying or oven heating prior to being frozen (Loewe, 1993). Critical battered food characteristics include viscosity, percentage of batter pickup, batter adhesion, oil absorption, appearance, and flavor. Generally, viscosity has direct implications on batter pickup. Batter pickup is regulated by the USDA and varies for product types; the amount permitted on meat or poultry products is 30% (USDA, 1997). Batter adhesion plays a critical role in battered food quality. Partial or total loss of coating during processing, frozen storage, transportation, and handling during consumption causes undesirable aesthetic and economic effects. Traditional deep-fat frying generates poor adhesion because the food shrinks away from the coating. Deep-fat frying, which is a common method for cooking battered products, can affect appearance, flavor, and oil

Abbreviation Key: HPMC = hydroxypropyl methylcellulose; MA = malonaldehyde; MC = methylcellulose; RFBB = rice flour-based batter; WFBB = wheat flour-based batter; WOF = warmed-over flavor; 6R0S0M = 60% rice flour:40% corn flour (wt/wt) and no oxidized corn starch or methycellulose; 6R15S0M = 60% rice flour:40% corn flour (wt/wt), 15% oxidized corn starch (dry basis), and no methylcellulose; 6R5S3M = 60% rice flour:40% corn flour (wt/wt), 5% oxidized corn starch, and 0.3% methylcellulose (dry basis); 5W0S0M = 50% wheat flour:50% corn flour (wt/wt) and no oxidized corn starch or methylcellulose.

2001 Poultry Science Association, Inc. Received for publication September 26, 2000. Accepted for publication March 13, 2001. 1 To whom correspondence should be addressed: therald@oznet. ksu.edu.

988

989

PHYSICOCHEMICAL PROPERTIES OF RICE FLOUR BATTER

and found that with appropriate levels of oxidized corn starch and methylcellulose (MC), RFBB exhibited good adhesion. This paper further elucidates the physicochemical and microbiological properties of RFBB on commercial chicken drumsticks.

MATERIALS AND METHODS Batter Ingredients and Formulations Batters were formulated with rice flour RL-1002 and yellow corn flour3 60:40 (flour weight) rice:corn flour; 0, 5, and 15% (dry basis) oxidized corn starch4; and 0 and 0.3% (dry basis) methylcellulose.5 All batter treatments were formulated with 5% salt, 2% sucrose, and 0.2% xanthan gum.6 To compensate for increases in oxidized starch and MC in the formulations, the percentage of flour was reduced while keeping the desired flour ratio constant. The solid to water ratio of batters was 1:1.3 (wt/wt). All dry ingredients were mixed at low speed for 1 min in a stainless-steel bowl in a Model K-45 mixer.7 Ingredients were then mixed with water for 2 min, cooled to 10 C in a refrigerator, and then stored in an ice bath to maintain the temperature during batter application.

Sample Preparation Individual quick frozen chicken drumsticks8 ranging in weight from 100 to 130 g/piece were thawed overnight in a refrigerator until their temperatures reached 4 C. Three separate batches of drumsticks consisting of two drumsticks per batch were prepared. Individual drumsticks were placed in a plastic bag9 and manually predusted with 1.3 g of egg white powder type P-11010 for 10 s. Drumsticks were dipped into batters and any excess batter was allowed to drip off for 10 s. Drumsticks were deep-fat fried in canola oil at 175 ± 5 C until the internal temperature reached between 71 to 75 C as measured with a meat thermometer at the thickest part of drumsticks. Then drumsticks were cooled to room temperature for 10 min prior to evaluation.

Storage Conditions After being fried, samples were held at room temperature for 10 min, placed in plastic bags11 (3 mil, nylon and

polyethylene bag, O2 transmission rate: 3.5 g/100 in2 per 24 h at 1 atm, 70 C, and 90% RH), and then sealed with packaging machine12 at speed number 1. Packed samples were kept in cardboard cartons and stored in an air blast freezer at −40 C for 90 d.

Composition Analysis Rice and corn flours were analyzed for starch damage using Megazyme kits13 following AACC Method 76.31 (AACC, 2000). Protein content was determined by nitrogen combustion analysis (AOAC, 1995a) with factor N × 5.95, 6.25, or 5.70 for rice, corn, and wheat flours, respectively, using Leco-FP-2000.14 Fat and moisture contents were determined by the rapid microwave-solvent extraction method using CEM15 (AOAC, 1995b).

Pickup and Cooking Loss Pickup refers to the amount of batter adhering to a food substrate during battering and is expressed as a percentage of the total product weight (Suderman, 1983). The modified method for pickup and cooking loss described by Olewnik and Kulp (1990) and Cunningham and Proctor (1984) were used, respectively. Drumsticks were dipped into batters, and excess batter was allowed to drip off for 10 s. The amount of batter pickup and cooking loss were calculated as follows:

cooking loss (%) =

Riviana Foods Inc., Houston, TX 77252. 3 ADM Milling Co., Lincoln, NE 68501. 4 National Starch and Chemical Co., Bridgewater, NJ 08807. 5 Dow Chemical Co., Midland, MI 48647. 6 Jungbunzlauer Inc., Newton Center, MA 02159. 7 KitchenAid Division, Hobart Co., Troy, OH 45374. 8 Tyson Foods, Inc., Springdale, AR 72765. 9 DowBrands L.P., Indianpolis, IN 46268. 10 Henningsen Foods Inc., Omaha, NE 68144. 11 Koch Supplies Inc., Kansas City, MO 64116. 12 Hollymatic Corporation, Countryside, IL 60525. 13 Megazyme International Ireland Ltd., Co., Wicklow, Ireland. 14 Leco Corporation, St. Joseph, MI 49085. 15 CEM Corp., Matthews, NC 28105. 16 Waring Products Division Dynamics Corporation of America, New Hartford, CT 06057.

[1]

(B − F) × 100 I

[2]

where B = battered chicken weight, I = initial chicken weight, and F = battered fried chicken weight.

Oil Absorption and Rancidity Analysis Fried battered chicken drumsticks were deboned and ground with liquid nitrogen for 30 s in a Waring blender.16 Ground samples were immediately analyzed for fat content using the CEM rapid microwave-solvent extraction method. The oil absorption was calculated as oil absorption (%) =

2

(B − I) × 100 I

batter pickup (%) =

Ffinal − Finitial × 100 Ffinal

[3]

where Finitial and Ffinal are fat contents in fried drumsticks before and after frying. The TBA test, which is expressed as milligrams of malonaldehyde (MA) per kilogram of sample, was performed as described by Witte et al. (1970). Frozen samples were thawed for 30 min before being tested.

Microstructural and Texture Analyses Microstructural and texture properties indicating adhesion of batter to drumsticks were evaluated using a laser

990

MUKPRASIRT ET AL. TABLE 1. Comparison between the physical properties of rice flour-based batters to traditional wheat flour-based batter at a fixed water to solid ratio (1.3:1, wt/wt)1 Property Batter pickup (%) Cooking lossNS (%) Fat content (%) Moisture content (%) Oil absorption (%)

6R0S0M a

30.00 18.82 10.48a 56.85b 49.95b

6R15S0M c

21.84 19.14 10.36a 57.52b 48.32b

6R5S3M b

26.14 18.57 9.20b 58.82a 46.40c

5W0S0M 24.58bc 19.21 10.53a 56.96b 50.57a

Means with different superscripts within the same row are significantly different (P < 0.05). Means within same row are not significantly different (P < 0.05). 1 6R0S0M = 60% rice flour:40% corn flour (wt/wt), 0% oxidized corn starch, and 0% methylcellulose; 6R15S0M = 60% rice flour:40% corn flour (wt/wt), 15% oxidized corn starch (dry basis), and 0% methylcellulose; 6R5S3M = 60% rice flour:40% corn flour (wt/wt), 5% oxidized corn starch (dry basis), and 0.3% methylcellulose (dry basis); 5W0S0M = 50% wheat flour:50% corn flour (wt/wt), 0% oxidized corn starch, and 0% methylcellulose. a–c NS

scanning microscope17 and a texture analyzer18 as described by Mukprasirt et al. (2000a).

Microbiological Analysis Three replications of fried battered chicken drumsticks with two drumsticks per replication were deboned and ground aseptically in a Waring blender. Microbiological testing was performed on 25-g samples from each treatment using the total plate count method on plate count agar.19 The colony-forming units were counted after incubation under aerobic conditions at 35 C for 48 h (Swanson et al., 1992).

Statistical Analysis All batter treatments were prepared and tested in three replicates with two subsamples per replication. Statistical analyses were performed using the general linear models procedure to determine the effect of storage time on physicochemical properties of frozen fried chicken drumsticks coated with RFBB and WFBB. The Bonferroni test was used to detect differences among means. Correlations between TBA values and fat and moisture contents, between moisture content and time, and between cooking loss and moisture content were analyzed using the Pearson correlation method. All statistical analyses were determined at a significance of P < 0.05 (SAS Institute, 1996).

RESULTS AND DISCUSSION Batter Pickup and Cooking Loss From our preliminary test, viscosities at 10 C of 6R0S0M, 6R15S0M, and 6R5S3M were similar to that of traditional wheat flour-based batter. Therefore, these formulations were selected for further study. Among RFBB, the 6R0S0M showed the highest pickup of 30%, whereas batter pickup of 6R15S0M was the lowest at 21.8% due to composition differences (Table 1). The amounts of rice

17

Carl Zeiss, Inc., Thornwood, NY 10594. Texture Technologies Co., Scarsdale, NY 10583. Difco Laboratories, Detroit, MI 48232.

18 19

and corn flours decreased as the level of modified corn starch increased in the ingredient formulation, which resulted in the 6R15S0M having less flour than other treatments. According to Cunningham and Tiede (1981), batter pickup is a function of batter viscosity. Mukprasirt et al. (2000b) found that RFBB containing higher flour levels exhibited greater viscosity because of damaged starch, which has a propensity to absorb water. The amounts of damaged starch in rice and corn flours were 7.03 and 2.43%, respectively. Our study was consistent with the results of Olewnik and Kulp (1990), who reported that the amount of batter pickup increased as flour protein level and damaged starch increased. By comparison, the amount of pickup for WFBB was not significantly different from that for the 6R5S3M or 6R15S0M but was lower (P < 0.05) than that for 6R0S0M. Gluten protein in wheat flour might have contributed to the batter viscoelastic property, which facilitated adherence to the drumsticks. Burge (1990) reported that batter pickup ranged between 24.4 to 39.6% when the ratios of corn to wheat flour were 0:1 and 2:1, respectively. The cooking losses of all batters were not significant (Table 1). Cunningham and Tiede (1981), however, found cooking loss to be substantially reduced with a more viscous batter and greater percentage of breading pickup, which may have been due to different batter formulations.

Fat and Moisture Contents Fat contents ranged between 9.20 and 10.53%, and moisture contents ranged between 56.85 to 58.82% (Table 2). The 6R5S3M treatment had a significantly lower fat content but a higher moisture content than other treatments. Perhaps a thermal gel formed by MC prevented mass transfer of moisture and oil during deep-fat frying. According to Meyers (1990), gums and starches serve as oil and moisture barriers more than other functional ingredients in batters. The MC and hydroxypropyl MC (HPMC) have been more widely investigated and used as oil barriers than any other hydrocolloids (Ang 1989; Lee and Han, 1988). Kuntz (1997) reported that these gums can reduce oil absorption up to 40% depending on factors such as frying time and temperature and surface to weight ratio. Balasubramaniam et al. (1997) reported


991

that the effectiveness of HPMC on fat reduction and moisture retention of a deep-fat fried poultry product were up to 16.4 and 33.7%, respectively, compared to a control.

Oil Absorption Percentage of oil absorption by fried chicken drumsticks differed (P < 0.05) as a result of batter formulation (Table 1). The 6R5S3M treatment exhibited the lowest oil absorption (46.4%) that may have been due to the addition of MC. Pinthus et al. (1993) reported that HPMC (Methocel K100M and F50LV) was more effective than powdered cellulose in reducing oil uptake of deep-fat fried donuts and falafel balls because of their thermal gelation and film-forming properties (Anon, 1980; Henderson, 1988). The 5W0S0M treatment exhibited the greatest oil absorption (50.57%) compared to other RFBB treatments. This increased absorption might have been due to differences in protein composition. Wheat gluten proteins might be responsible for oil absorption in fried batter. Gluten protein can expand during deep-fat frying because of an intrinsic viscoelastic property, resulting in a fluffy coating that may facilitate water and fat transfer. Water loss from food being deep-fat-fried lowers the internal pressure, allowing penetration of the frying medium. Conversely, the RFBB might not have expanded because of a gluten deficiency, resulting in a different product surface and shape compared to WFBB. Annapure et al. (1998) stated that surface properties rather than chemical composition or physicochemical properties were primarily responsible for oil uptake during deep-fat frying. Their results showed that wheat flour had a higher oil-holding capacity than rice flour, whereas rice flour had a higher TABLE 2. Peak force (Newtons) needed to pull selected batters off fried chicken drumsticks using an attachment developed for texture analyzer Storage time1 Treatments2

Day 0

Day 30NS

Day 60

Day 90NS

6R0S0M

8.1dx (0.5) 10.9bx (0.9) 11.6ax (1.1)

7.6x (0.7) 8.2y (0.8) 8.5y (0.4)

6.5aby (0.4) 7.4ayz (0.6) 6.3bz (0.5)

5.7y (0.8) 6.5z (0.7) 6.2z (0.7)

9.8cx (0.5)

8.5y (0.7)

7.4ay (0.7)

6.1z (0.6)

6R15S0M 6R5S3M 5W0S0M

a–d Means (SD) of three replications with different superscripts within the same column were significantly different (P < 0.05). x–z Means (SD) of three replications with different superscripts within the same row were significantly different (P < 0.05). NS Means (SD) of three replications with different superscripts within the same column were not significantly different (P < 0.05). 1 Storage period at −40 C. 2 6R0S0M = 60% rice flour:40% corn flour (wt/wt), 0% oxidized corn starch, and 0% methylcellulose; 6R15S0M = 60% rice flour:40% corn flour (wt/wt), 15% oxidized corn starch (dry basis), and 0% methylcellulose; 6R5S3M = 60% rice flour:40% corn flour (wt/wt), 5% oxidized corn starch (dry basis), and 0.3% methylcellulose (dry basis); 5W0S0M = 50% wheat flour:50% corn flour (wt/wt), 0% oxidized corn starch, and 0% methylcellulose.

FIGURE 1. The TBA values of fried chicken drumsticks coated with selected rice flour-based batters compared to wheat flour-based batter during 90 d of frozen storage; 6R0S0M = 60:40% (flour wt) rice:corn flours without oxidized corn starch or methylcellulose; 6R15S0M = 60:40 rice:corn flours with 15% (dry basis) oxidized corn starch; 6R5S3M = 60:40 rice:corn flours with 5% oxidized corn starch and 0.3% (dry basis) methylcellulose; 5W0S0M = 50:50% (flour wt) wheat:corn flours batter; MA = malonaldehyde.

water-holding capacity than wheat flour at 30 to 80 C. Additionally, differences in chemical structure of rice and wheat proteins might play a critical role in oil absorption. Our results agreed with Shih and Daigle (1999) who reported that oil retention in RFBB and WFBB were 27.6 and 49.3%, respectively. They explained that wheat protein binds tightly with oil molecules, thus increasing fat content; however, the mechanism is not clear. Olewnik and Kulp (1990) reported that chicken drumsticks coated with wheat flour (7 to 12% protein) absorbed approximately 49 to 64% oil depending upon flour protein content. Batter from higher-protein flour yielded more fat absorption but less moisture retention. In contrast, the authors found that the level of flour protein did not influence the perceived greasiness of fried coating evaluated by subjective scoring. Baker and Scott-Kline (1988) found that batters containing pregelatinized or modified flour produced coatings with high moisture content and absorbed little fat. The breading reduced the perception of rubberiness and greasiness. Makinson et al. (1987) reported that oil absorption in plant foods with high initial water content and low fat content was higher than that in animal foods. Battered fish without breading markedly retarded oil absorption because of the rapid formation of a hard crust, which was relatively impervious to the movement of water and fat.

TBA Analysis The TBA values of frozen prefried battered chicken drumsticks increased (P < 0.05) with storage time (Figure 1). The relationship between time and TBA values was positive with a correlation coefficient (r) of 0.76. The TBA value of treatment 6R5S3M was the lowest compared to other treatments at any period, possibly because MC formed a thermal gel during frying. This gel would act as

992

MUKPRASIRT ET AL.

an oil barrier (Meyers, 1990), thus reducing oil absorption. The TBA values of WFBB were significantly higher than those of RFBB treatments after 60 d of frozen storage, but all values were in the same ranges reported by other researchers. Wang et al. (1976a) reported that TBA, peroxide, and acid values of frozen prefried chicken products from a grocery store ranged between 2.1 to 9.2 mg MA/ kg, 7.0 to 25.5 meq/kg, and 0.99 to 2.64 mg KOH/g, respectively. Although the rancidity indicators were high, they reported that no rancid odor was detected. Igene et al. (1985) investigated TBA-reactive substances in relation to warmed-over flavor (WOF) development in cooked chicken. They reported that chicken with TBA values of 2.07 to 2.41 mg MA/kg was described as having pronounced WOF by panelists. Ang and Huang (1993) reported TBA values from 2.59 to 3.47 mg MA/kg for chicken patties. The average TBA value for all treatments was lower (P < 0.05) at Day 0 compared to other storage periods. TBA increased as the storage time increased except at Days 7 and 15, which were not significantly different from each other. The mean TBA values at Day 0 was 1.4 mg MA/kg sample, whereas it increased to 2.77 mg MA/kg sample at Day 90. The TBA of treatments 6R15S0M and 6R5S3M decreased from Days 7 to day 15 and then increased through the remaining storage period. The reason for this observation might be the nature of the TBA analysis, because short-chain carbon products of lipid oxidation are not stable (Fernandez et al., 1997). The oxidation of these products results in organic alcohols and acids, which are not determined by the TBA test (Tarladgis and Watts, 1960; Seo, 1976; Almandos et al., 1986). Additionally, MA can react with amino acids thus decreasing TBA values (Buttkus, 1976; Gardner, 1979). However, Fernandez et al. (1997) reported that the TBA test was the most sensitive to detect linolenic and linoleic acid oxidation products. Lai et al. (1991) found that the TBA values of control frozen chicken nuggets kept up to 6 mo were 1.67 to 3.64 mg MA/kg and decreased after 4 mo of frozen storage. Tomas and Anon (1990) found that storage time significantly affected TBA values in salmon and chicken breast muscles, but the freezing rate did not. The relationship between fat content and TBA values in our study was positive (r = 0.73). These results suggested that increasing fat content promoted increased lipid oxidation and resulted in higher TBA values. This observation was expected, because polyunsaturated fatty acids accounted for most of the lipid composition in these samples. The abundance of polyunsaturated fatty acids, which are TBAR-reactive substance precursors, in chicken meat contributed to oxidation susceptibility (Melton, 1983; Pikul et al., 1985). In addition, NaCl used in the formulation might have had an effect. Kanner et al. (1991) reported that NaCl promoted lipid oxidation by enhancing iron ion activity. An adverse effect of NaCl was noted in the presence of air and absence of antioxidants (Jul, 1984). Method of freezing was another factor affecting lipid oxidation. Berry and Cunningham (1970) reported results from a TBA test suggesting that freezing rate af-

fected the degree of rancidity in frozen fried chicken. Liquid nitrogen was the most desirable method to reduce lipid oxidation during freezing, and use of a household freezer was the least desirable. Although some treatments showed high TBA values, rancid odor was not detected in this study by authors in an informal panel. Lai et al. (1995) reported that the correlation coefficient between TBA values and sensory score was low. The TBA test may be less sensitive to monitoring WOF in meats in long-term frozen storage (after 4 mo), because of the instability and reactivity of MA. However, the TBA test appeared to detect the variation over storage time better than sensory evaluation. Greene and Cumuze (1981) reported the relationship between TBA values and inexperienced panelists’ assessments of oxidized flavor in cooked beef. They found that panelists initially detected oxidized flavor when the TBA values ranged from 0.6 to 2.0 mg MA/kg, whereas Tarladgis et al. (1960) found that trained panelists recognized rancid odor at 0.5 to 1.0 mg MA/kg. Melton (1983) found that rancid flavor was detectable at 0.3 to 1.0 mg MA/ kg in beef or pork, greater than 3.0 mg MA/kg in turkey, and 1.0 or 2.0 mg MA/kg in chicken. However, these ranges should not be used as general references for thresholds of rancid flavor in meat, because TBA values are affected by dietary status, age of animals prior to slaughter, and methods used for TBA analyses (Fernandez et al., 1997). A threshold for detectable rancidity in fried chicken products is needed.

Microstructural Analysis Micrographs of battered fried chicken drumsticks comparing RFBB formulations to WFBB samples at 0, 30, 60, and 90 d are shown in Figures 2 to 5, respectively. At 0, 30, 60, and 90 d, all samples with different batter formulations showed good adhesion between batter and chicken skin. Differences between treatments within the batter layers were observed over the 90 d frozen storage. Initial observations indicated voids (Figure 2) within the batter layer. Voids observed at 0 d may be from water evaporated or be the result of batter expansion during frying. These voids were smallest in size and less in numbers in the wheat formulation as compared to the larger more numerous voids in the 6R5S3M formulation. One possible explanation for this observation may be that the batter containing MC had a higher moisture content. The MC had a high water absorption (Glickman, 1969) and has been reported to form a thermal gel (Dow Chemical, 1996) that may prevent water evaporation during frying. Number and size of voids increased within the fried batter during storage due to water retention, which upon slow freezing would lead to ice crystal growth. Micrographs at 90 d of frozen storage (Figure 5) revealed that the 6R5S3M (Figure 5-D) treatment exhibited more damage compared to other treatments. Our results were similar to findings of Corey et al. (1987), who reported that differences in the batter ingredients and freeze-thaw cycle did not significantly affect adhesion of breaded fish portions,


993

FIGURE 2. Microstructure of fried chicken drumsticks coated with (A) wheat flour-based batter; (B) 6R0S0M, rice flour-based batters, 60:40% (flour wt) rice:corn flours; (C) 6R15S0M, 60:40% rice:corn flours with 15% (dry basis) oxidized corn starch; and (D) 6R5S3M, 60:40% rice:corn flours with 5% oxidized corn starch and 0.3% (dry basis) methylcellulose at 0 d, as visualized by laser scanning confocal microscopy. Arrowheads indicate voids that possibly occurred during frying or frozen storage.



FIGURE 5. Microstructure of fried chicken drumsticks coated with (A) wheat flour-based batter; (B) 6R0M0S, rice flour-based batters, 60:40% (flour wt) rice:corn flours; (C) 6R15S0M, 60:40% rice:corn flours with 15% (dry basis) oxidized corn starch; and (D) 6R5S3M, 60:40% rice:corn flours with 5% oxidized corn starch and 0.3% (dry basis) methylcellulose at 90 d, as visualized by laser scanning confocal microscopy. Arrowheads indicate voids that possibly occurred during frying or frozen storage.

994

MUKPRASIRT ET AL. TABLE 3. Total plate counts (log cfu/g) of raw materials and frozen fried chicken drumsticks Storage time2 Treatments1

Day 0

Day 30

Day 60

Day 90

Raw drumsticks Raw batter 6R0S0M 6R15S0M 6R5S3M 5W0S0M Fried battered drumsticks 6R0S0M 6R15S0M 6R5S3M 5W0S0M

5.78

...

...

...

3.61 3.26 3.50 3.48

. . . .

. . . .

. . . .

. . . .

. . . .