Influence of Outer Layer Formulation on the Sensory Properties of Microwaved Breaded Nuggets

Abstract

The present study analyzed the effect of the type of first layer (egg or commercial batter) and the number of coatings (one or two) on the final sensory quality of microwaved breaded products (chicken nuggets). The products were evaluated by sensory and instrumental methods. Sensory analysis was carried out by a panel of eight panelists trained in descriptive analysis of battered and breaded products. The instrumental parameters measured were color, texture and sound emission, and moisture and fat contents. A two-way ANOVA was performed on the sensory and instrumental results and Fisher's least significant differences were calculated. Sensory assessment was more suitable for evaluating microwaved breaded foods, such as chicken nuggets.

Keywords:

• Breaded nuggets
• Egg
• Commercial batter
• Microwave
• Sensory properties
• Mechanical properties

INTRODUCTION

There is no doubt that battered and breaded foods are highly appreciated and very widespread throughout the world. The numerous examples in different countries include fish and chips in England, fried chicken in the USA, milanesas (breaded veal) in Argentina, tempura in Japan, and pescaito frito (fried fish) in Spain, among many others. A large food service sector revolves around battered and breaded foods, mainly fast food restaurants. These would benefit from the possibility of having pre-fried breaded food in a freezer and using a microwave for the final cooking/heating when a client asks for this product, thus saving time both for the restaurant and for the client. During frying or oven baking, the outer coating is the part that is most exposed to high temperatures (hot oil in the fryer or hot air in the oven), which dehydrates it and thus provides a crunchy crust that is a key sensory quality factor. Breaded products present a sandwich-like structure, with layers of different composition, particularly regarding the content of plasticizers, such as water or oil. What is desired in these products is a soft and deformable interior surrounded by a dry, firm, and crunchy crust. This final structure is highly dependent on the compositions of core and crust after processing: water content, size of the food piece, and composition of the coating material. It also depends on processing variables: cooking procedure, cooking time, and temperature. Microwaved products usually tend to lack the desired crunchiness that is typical of a fried breaded food. The main reason is that in microwave heating the food pieces begin to be heated from the center. Moisture in the food piece is driven outwards, as vapor, towards the surface, where it condenses because the temperature inside the microwave oven is below 100°C, causing the food piece to become soggy. Thus, it would be interesting to investigate whether there is a processing factor or a combination of processing factors that could improve the sensory quality of the final product.

The latest trends indicate that a combination of the appropriate packaging technology and a balanced choice of food formulation ingredients could be a partial solution to the problem of quality loss through microwave heating of frozen pre-fried battered and breaded food. Certain patented procedures/products claim that they improve the crunchiness of microwaved battered foods by using starchy products, such as high-amylose flour[1,2] or genetically modified potato starch[3] as one of the batter ingredients.[4] The influence of dextrins of two different chain-lengths and the use of a susceptor material to improve the sensory crunchiness of microwaved battered fish have been studied. These published articles usually refer to tempura-battered foods; a lower number report research into breaded foods.

One of the most popular breaded foods is nuggets. The structure of the breaded chicken nugget crust is composed of two layers. One is an adhesion batter, usually beaten egg or a starchy slurry preparation, and the other is the outer layer of bread crumbs, corn meal, cracker crumbs, or the like. There are no studies of the effects of changing the features of each of these two coverings on the crunchiness of microwaved products. Obviously, the final quality would have to be defined by sensory analysis, but instrumental techniques could be an interesting complement.

Traditionally, home-made breaded products use a first layer of beaten eggs (before the bread crumbs are applied). In food service systems, for health reasons, the regulations usually only allow pasteurized eggs. The handling and short shelf-life of pasteurized egg can present problems in fast food restaurant facilities. Starchy preparations, which can be reconstituted on the go, are an attractive alternative from the practical and safety points of view. Consequently, analyzing the effect of both types of first layer on the quality of the microwaved product is of interest. When preparing breaded products, the piece of food (veal, chicken, cheese, or others) is dipped in the beaten egg and then covered with bread crumbs. In home and restaurant cooking it is a regular procedure to perform this coating process once (egg-breadcrumbs) or twice (egg-breadcrumbs-egg-breadcrumbs); this second procedure is a way of achieving greater structural strength and also, obviously, greater yields. The choice of one or two coatings can also be made for sensory reasons.[5] Hydroxypropyl methylcellulose (HPMC) was used to act as a barrier to oil absorption and moisture loss with the hope of maintaining crispiness and eating quality in microwaved battered products and some improvements were obtained. It could be hypothesized that having two coatings could act as a natural barrier (or at least an obstacle) before the food moisture reaches the external covering and thus prevents this moisture from decreasing the crunchiness of the final product.

Despite the fact there are numerous works that correlate sensory and instrumental measurements of complex foods,[6,7] there is an interest in searching for correlations between sensory and instrumental measurements for microwaved breaded products.[8] Researchers studied the correlation between instrumental and sensory measurements of crispness for chicken nuggets, finding a positive correlation between mechanical Kramer shear-compression cell parameters, non-destructive ultrasonic pulse measurements parameters, and sensory crispness measured by a trained panel.[4] It was found that the instrumental parameters were correlated with trained panel descriptors for fish nuggets with a dextrin-added batter formulation.

The main objective of the present work was to analyze the influence of different first layers (egg or batter) and number of coatings (one or two) on the final sensory quality of microwaved breaded nuggets. A complementary objective for these structurally complex products was to analyze the relationship between sensory and instrumental texture, color, and composition parameters.

MATERIALS AND METHODS

Sample Preparation

Nuggets

Fresh boneless minced chicken breasts were purchased from a local shop. The meat pieces were blended with a commercial seasoning for nuggets in a Kenwood Major Classic mixer (Kenwood Ltd., Watorford, UK); the seasoning contained salt, starch, spices, dextrose, E-301, E-331, and coloring; 40 g was added per 1 kg of chicken. Using a mold, rectangular chicken blocks (nuggets) measuring 4 × 3 × 1 cm were formed and stored in a freezer at −18°C until used.

Coating

Preformed nuggets were taken out of the freezer and left at room temperature for 10 min. After this time they were coated according to the following 2×2 factorial design:

•	Type of first layer: Beaten total pasteurized egg (Ovocity, Valencia, 106 Spain) or a commercial batter (Adinmix, composed of wheat flour, pregelatinized corn starch, salt, and guar gum, with a 1:6 solid to water ratio).
•	Number of coatings: One (first layer/breading) or two (first layer/breading/first layer/breading).

For the first layer, the preformed nuggets were dipped for 10 s in the egg or commercial batter and then the excess was removed. Then they were breaded with a fine particle size breading (type TN-M30) for 5 s for each side. The commercial batter and breading were provided by Adín, S.A. (Paterna, Spain).

Prefrying

The nuggets were prefried in sunflower oil at 190°C for 30 s (15 s on each side) in a Fritaurus Professional 4 domestic fryer (Taurus, Barcelona, Spain), left to cool and placed in plastic freezer bags (LDPE film, thickness 150 microns), stored at -18°C for a week, and then cooked without thawing. The pre-frying step gave the nuggets a preliminary golden-brown color and partially cooked them. The final cooking was carried out in a domestic microwave oven (Samsung M1727, Barcelona, Spain), 1 nugget at a time for 50 s at maximum power (900 Watts).

Sample Analysis

The samples with egg were coded with the prefix ‘E’ followed by 1 or 2 depending on the number of coatings, giving E1 and E2. In the same way, the samples dipped in the prepared batter were coded with the prefix ‘B’ and the number of coatings, giving B1 and B2. Two replicates were prepared on different days and four nuggets per treatment (E1, E2, B1, and B2) were determined in each analysis.

Table 1  Sensory analysis descriptors and evaluation techniques

Descriptor	Description	Technique
Golden brown color	Typical golden brown color of fried products	Observe whole surface of nugget
Crunchiness when cutting	A measure of the sound emitted when cutting the nugget	Cut the nugget in half with a knife
Hardness when cutting	A measure of the necessary force to cut the nugget	Cut the nugget in half with a knife
Ease of cutting	A measure of the resistance of the nugget while it is being cut	Cut the nugget in half with a knife
Adherence	Amount of crust that remains stuck along the chicken piece perimeter after cutting	Observe the transversal surfaces of the two nugget halves
Visual juiciness	A measure of the surface appearance of moistness	Observe the transversal surfaces of the two nugget halves
Layer distinctness	A measure of the distinction of the different layers along chicken piece perimeter	Observe the transversal surfaces of the two nugget halves
Crust thickness	An estimated measure of the crust thickness	Observe the transversal surfaces of the two nugget halves
Crunchiness	A measure of the sound emitted while chewing	Bite a nugget in half
Hardness	A measure of the force while biting	Bite a nugget in half
Juiciness	A measure of the rate of liquid release while biting	Bite a nugget in half

Sensory Assessment

An 8-member quantitative descriptive analysis (QDA®)[9] panel drawn from IATA (Institute of Agrochemistry and Food Technology) personnel was trained and used to evaluate several attributes of the nuggets. The whole panel had one-year experience in nugget quality evaluation, and the panelists were members of a group being trained to function as a permanent descriptive analysis panel for a variety of foods. The selection of descriptors was made over two 1-h sessions. The panelists were provided with representative samples and asked to choose the words that best described them.[10] As well as the samples with final cooking in a microwave oven, they were also presented with samples with a fried final cooking stage so that the product most often encountered in the market was available for reference. During the training sessions, the panelists suggested a list of meaningful sensory attributes for the samples and discussed the definition and how to evaluate each one.[9] The panelists took part in eleven 1-h training sessions over a period of 2 months. Once the terms had been selected, consensus concerning their use was attained; this entailed reaching a precise definition of the descriptors and how to evaluate them to quantify attribute intensity, as well as agreeing upon the testing procedure. The trained panel established 11 descriptors with their corresponding descriptions and measuring techniques (Table 1). The trained sensory panelist's performance was monitored using general Procrustes analysis and by analyzing panelist × sample interactions for each descriptor. The discriminative power of each individual panelist was assessed using an analysis of variance (ANOVA). A balanced complete block experimental design was carried out to evaluate the samples. Each assessor scored the nuggets for each term using 10-cm unstructured scales, anchored on the left end with “low” intensity and on the right end with “high” intensity. The samples were served on a separate plastic tray identified with a random three digit code. Water and green apple pieces were provided to panelists during the sessions to minimize the residual effect between samples.[11]

Instrumental Analysis

Color

Color was measured instrumentally with a Konica Minolta CM-3500d spectrophotometer (Konica Minolta Sensing Inc., Osaka, Japan). The results were expressed in accordance with the CIELAB system, with reference to illuminant D65 with a visual angle of 10°. The measurements were performed through a 8-mm diameter diaphragm containing an optical glass. The CIELAB parameters (L*, a*, b*, Chroma-C*_ab, and hue-h_ab) for all samples were determined following the recommendations of the Commission Internationale de L'Eclairage.[12] Color differences (ΔE*_ab), which are important for evaluating relationships between visual and numerical analyses, were calculated as the Euclidean distance between two points in the three-dimensional space defined by L*, a*, and b*:

(1)

All pairs of samples were compared (E1-E2, B1-B2, E1-B1, and E2-B2).

Texture and sound emission

A TA-XT plus Texture Analyser (Stable Micro Systems, Godalming, UK) with a 25 kg load was used for cutting tests. Sample cutting was performed with a light knife blade (A/LKB). The samples were placed on the HDP/90 Heavy Duty Platform with a slotted blade insert, with the larger side of the nugget placed perpendicular to the blade. The test settings were test speed 1 mm/s, trigger force 5 g, and blade displacement 20 mm (in order to cut completely through the sample). An acoustic envelope detector (AED) was used for sound recording during the cutting tests; the experimental conditions were adapted from Varela et al.[13] The gain of the AED was set at one. A Bruel and Kjaer free-field microphone (8-mm diameter), calibrated using a Type 4231 Acoustic Calibrator (94 dB and 114 dB SPL-1000 Hz) (Bruel & Kjaer, Naerum, Denmark), was placed in a frontal position in order to gain a better acoustic signal, at a distance of 4 cm and an angle of 45° to the sample. A built-in low pass (anti-aliasing) filter set the upper calibrated and measured frequency at 16 kHz. Ambient acoustic and mechanical noise was filtered by the use of a 1 kHz high pass filter. The AED operates by integrating all the frequencies within the band pass range, generating a voltage proportional to the sound pressure level (SPL). The data acquisition rate was 500 points per second for both force and acoustic signals. All the tests were performed in a laboratory with no special soundproofing facilities at an ambient temperature of 22 ± 2°C. Force as a function of displacement and SPL as a function of displacement were plotted simultaneously. The parameters extracted from the curves were:

•	Number of force peaks counted using a threshold value of 0.0981 N;
•	Number of sound peaks counted using a threshold value of 2.5 dB;
•	Max SPL (dB): Maximum peak intensity of the sound pressure level (SPL), which is a measure of the sound level or the loudness of the sound events.

Mean values of the number of force peaks, number of sound peaks, and maximum SPL have proved to be good indicators of the crispy/crunchy character of food items.[14–16 16 Data management was performed with Texture Exponent 32 software (Stable Micro Systems, Godalming, UK).

Moisture and total fat contents

After cooking and cooling to room temperature, the crust and core were separated to analyze the moisture and fat contents of each part. Fat extraction was determined by adapting the 991.36 AOAC official method[17] for fat in meat and meat products. The fat was extracted using a FOSS Soxtec Avanti 2055 manual system (Tecator, Höganäs, Sweden) based on continuous extraction with petroleum ether for 3 h and 10 min.

The crust and chicken core of the samples were homogenized separately (Moulinex Food Mincer, Barcelona, Spain) and about 1.5 g were weighed out into extraction thimbles housed in dried, weighed metal containers. Sea-sand was also added to ensure the smooth extraction of the fat. Eighty grams of petroleum ether were used as the solvent for each extraction and the thimble contents were heated to a temperature of 135°C to extract the fat. The extraction times used were as follows: boiling 60 min, rinsing 80 min, and recovery 50 min. When extraction was finalized, the metal containers containing the extracted fat were placed in a heater (100°C/1 h) to ensure that all the solvent evaporated, then in a desiccator for 15 min to cool down sufficiently for accurate weighing. An electronic balance was used to measure the weight of the metal containers. The formula: %Fat = (W1 - W2)/W3 * 100% was used to calculate the percentage of fat in the dry weight of the sample. W1 is the weight of the metal container after the extraction, W2 is the weight of the metal container before the extraction, and W3 is the weight of the dry, homogenized samples weighed out into the thimbles. The result of the division is then multiplied by 100% to obtain the percentage of fat (dry basis). The moisture was determined by vacuum drying (10⁻² mmHg) at 95°C to a constant weight following AOAC method 950.46.[17]

Data Analysis

A two-way ANOVA (layer type, number of coatings, and the layer type * number of coatings interaction) was performed on the sensory and instrumental results. For the sensory data, the assessors were considered as an additional random effect. Fisher's Least Significant Differences were calculated for each term. Linear Partial Least Squares Regression Analysis (PLS) was used to analyze the relationships between sensory and physical matrices.[18] PLS extracts a few linear combinations (PLS factors) of the physical data that predict as many of the systematic variations in the sensory data as possible. PLS is recommended when there is a reduced number of samples in relation to the number of parameters to be correlated, and also when there is collinearity between the parameters as expected in the present analysis. Osten's F-test[19] was used to determine the number of significant (P ≤ 0.05) factors. This test compares the cross-validation sum of squares from the current dimension with the change of cross-validation sum of squares from the previous dimension. Data analyses were performed using Genstat 13th Edition (VSN International, Rothamstead, UK).

RESULTS

Sensory Analysis

Table 2 summarizes F and p-values of the ANOVA analysis of sensory and instrumental measurements. In the generation of descriptors, particular attention was paid to those related to texture, given the importance of this attribute for characterizing products with a sandwich structure (moist and juicy inside and dry and crunchy on the outside). The only descriptors that showed significant interactions were ‘adherence’ (Fig. 1a) and ‘visual juiciness’ (Fig. 1b). In the samples prepared with egg, only the double coating method caused a reduction in crust adherence; in the case of visual juiciness only the double-coated commercial batter samples presented increased juiciness. For descriptors, such as ‘golden brown’, ‘crunchiness when cutting’, ‘ease of cutting’, and ‘crunchiness’, the panelists found no significant differences in intensity between first layer type and number of coatings (Table 3).

Table 2   F- and p-values from analysis of variance for the outer-layer formulation factors: type of the first layer, number of coatings and their interaction, for sensory and instrumental measurements

Measurements	Type of first layer		Number of coatings		Two-way interaction
Sensory assessment	F	p-value	F	p-value	F	p-value
Golden brown color	0.1	0.79	3.1	0.10	0.0	0.85
Crunchiness when cutting	0.0	0.90	1.7	0.21	0.1	0.75
Hardness when cutting	5.4	0.04	3.3	0.10	1.0	0.34
Ease of cutting	0.7	0.43	0.8	0.38	0.0	0.96
Adherence	8.5	0.01	7.4	0.02	8.3	0.01
Visual juiciness	13.7	0.00	1.5	0.25	1.6	0.04
Batter layer distinctiveness	5.4	0.04	27.6	0.00	0.7	0.43
Crust thickness	1.4	0.26	32.2	0.00	1.3	0.28
Crunchiness	0.6	0.45	0.9	0.35	0.0	0.94
Hardness	21.3	0.00	4.1	0.07	0.0	0.95
Juiciness	46.4	<0.0001	5.4	0.04	0.0	0.83
Color
L*	3.1	0.11	0.2	0.69	3.3	0.09
C	1.2	0.29	0.4	0.54	0.1	0.74
H	1.0	0.33	1.1	0.31	0.6	0.47
Texture and sound emission
Number of sound peaks	13.9	0.00	2.8	0.12	22.2	0.00
Number of sound peaks	2.1	0.17	11.7	0.01	0.0	0.94
Max SPL	1.9	0.20	1.5	0.24	0.5	0.49
Moisture and fat content
Crust moisture	32.9	<0.0001	3.7	0.08	0.1	0.71
Crust fat	188.7	<0.0001	23.1	0.00	12.4	0.00
Core moisture	25.8	0.00	18.9	0.00	15.7	0.00
Core fat	18.0	0.00	5.6	0.04	0.2	0.65

Table 3  Mean and standard deviation of sensory QDA values for samples with one coating of commercial batter and breading (B1) or two (B2), or with one coating of beaten egg and breading (E1) or two (E2)

Attribute	B1	B2	E1	E2
Golden brown color	5.9 ± 0.6^a	6.6 ± 0.8^a	6.0 ± 1.1^a	6.8 ± 0.7^a
Crunchiness when cutting	3.0 ± 0.2^a	3.8 ± 0.8^a	3.2 ± 1.6^a	3.7 ± 1.0^a
Hardness when cutting	6.1 ± 0.4^ab	4.9 ± 0.5^a	6.6 ± 1.4^b	6.3 ± 0.6^a
Ease of cutting	5.7 ± 0.7^a	6.0 ± 0.3^a	5.4 ± 1.2^a	5.7 ± 0.4^a
Adherence	6.8 ± 0.8^a	6.9 ± 0.8^a	6.8 ± 0.6^a	3.9 ± 1.6^b
Visual juiciness	4.9 ± 0.5^ab	5.6 ± 0.3^a	4.2 ± 0.6^b	4.2 ± 0.8^b
Layer distinctiveness	4.5 ± 0.7^ab	6.7 ± 1.5^a	2.9 ± 0.4^b	5.9 ± 1.0^c
Crust thickness	3.5 ± 0.5^a	6.2 ± 1.8^b	2.2 ± 0.3^a	6.2 ± 1.4^b
Crunchiness	3.1 ± 0.4^a	3.5 ± 0.6^a	2.6 ± 1.5^a	3.2 ± 1.3^a
Hardness	4.9 ± 1.1^ab	4.0 ± 1.3^a	7.2 ± 0.3^c	6.2 ± 0.9^bc
Juiciness	5.9 ± 0.7^a	7.0 ± 0.7^a	3.1 ± 0.1^b	4.0 ± 1.4^b
Identical letters in the same row indicate that there is no significant difference at p > 0.05 (Fisher´s least significant differences).

Figure 1 (a) Interaction between number of coatings and type of first layer (egg or batter) for sensory adherence. (b) Interaction between number of coatings and type of first layer (egg or batter) for sensory visual juiciness.

The samples prepared with the commercial batter, whether single- or double-coated (B1 and B2), were found to be less hard than those prepared with egg, both during manual cutting and in the mouth, and also juicier. This could be explained by the presence of hydrocolloids, such as the pregelatinized corn starch and guar gum, that were contained in the batter formulation, as these help to retain water and give a moist, juicy mouthfeel. The ‘number of coatings’ effect was reflected by significant differences in the ‘crust thickness’ values; in the double-coated samples, the crust was perceived as being thicker, as would be expected. For ‘crunchiness when cutting’ and ‘crunchiness’, the double-coated samples showed a tendency to be slightly crunchier than those with a single coating, although the data were not statistically significant. It must be pointed out that, in general, the values registered for both crunchiness descriptors were low (around 3) because all the study samples were microwaved, whereas fried product samples (with a value of 10) were used for reference purposes during the panel training. Despite the absence of significant differences, this tendency to find the double-coated samples crunchier is understandable, bearing in mind that as the number of layers increases, the moisture which is transferred from the interior to the exterior of the nugget in the microwave oven has to travel a greater distance and may not reach the condensation point on the surface, which will therefore remain crunchier. Also, the presence in the mouth of more pieces of nugget with breadcrumb particles will help the perception of crunchiness to continue for longer, increasing the perceived intensity of this attribute.

Color

The ANOVA results for the instrumental color data show no significant differences in saturation (C*_ab) or hue (h_ab) between the samples, as all the values were around 32.5 and 67.5, respectively. These values indicate a yellower and less saturated color than in the fried samples, which also agrees with the sensory analysis results for the ‘golden-brown’ descriptor, where the microwaved samples scored around 6–7 against 10 for the fried reference samples in the training period. The E1 samples were the darkest, with L* values of 41.8 compared to a mean L* of near 36 for the rest, among which there were no significant differences. A significant interaction was also found, in that the samples prepared with the commercial batter in a double coating were as dark as those with a single coating, whereas the samples prepared with egg in a single coating were darker than those with a double coating. While the study samples were pre-fried for 30 s, this contributed little goldenness, and they did not reach temperatures in the microwave oven that would assist darkening reactions; as a result, these samples were paler than the fried ones and their surfaces displayed few nuances of color.

The human eye is considered capable of perceiving color differences when the ΔE* value is greater than 3.[20] The sample pairs that exceeded this ΔE* value were E1-B1 and E1-E2 (ΔE* = 6.5 and 3.8, respectively), a result which is principally attributed to the difference in lightness of the samples.The panelists' scores did not find significant differences in golden brown intensity due to layer type or number of coatings. It should be mentioned that in the instrumental color determinations, the only significant difference was in luminosity, which was not the descriptor that the panel chose for characterizing this type of sample.

Mechanical Measurements and Sound Emission

The typical profile of a fried nugget had numerous force peaks and a jagged plateau shape (Fig. 2), whereas in the microwaved sample profiles of the force peaks disappeared and practically no fracture event was seen because the samples were not crunchy. The number of force peaks and number of sound peaks are related to greater crunchiness, although it must be borne in mind that for this type of product not all sound events correspond to fracture events, since it was found that factors, such as external oil bubbles or air pockets, can also make a sound when they break but do not contribute to the sensory perception of crunchiness with the same intensity as do fracture events when biting.

Figure 2 Typical profile of a fried nugget. Force (black line) and sound pressure level (SPL, grey line) versus distance for chicken nuggets cooked by deep frying.

The instrumental texture of samples E1, E2, B1, and B2 (Table 4) show the profile of a nugget with low crunchiness values with a low number of peaks during cutting. This number of force peaks was somewhat higher to values obtained by other authors[13] that have worked with commercial chicken nuggets. This may be because many commercial nuggets have a tempura-type covering, in other words, they do not have an external layer of breadcrumbs that would cause a greater number of fracture events when broken by cutting. Samples with double coating presented higher values for number of force and number of sound peaks (more crunchy character). In the sensory evaluation the ‘crunchiness when cutting’ and ‘crunchiness’ descriptors showed the same tendency: the double-coated samples were perceived as being crunchier. Double coating confers greater protection of the nugget's internal juices, as well as a greater quantity of breadcrumb particles. Nevertheless, the samples with a single and double coating presented differences in the number of sound peaks whereas the sound level values were similar, so for this kind of sample, where only subtle differences existed, the latter sound parameter was not a satisfactory indicator of crunchiness.

Table 4  Instrumental determinations for samples with one coating of commercial batter and breading (B1) or two (B2), or with one coating of beaten egg and breading (E1) or two (E2)

	Blade cutting
	Mechanical	Sound
Sample	No. of force peaks	No. of sound peaks	Max. SPL (dB)	Crust moisture (g/100 g) db	Crust fat (g/100 g) db
B1	6.3 ± 2.2^a	4.0 ± 2.2^a	61.3 ± 1.7^a	40.5 ± 2.7^a	15.7 ± 1.4^a
B2	18.3 ± 4.1^b	9.8 ± 4.5^bc	61.0 ± 1.3^a	42.2 ± 2.5^a	17.0 ± 1.9^a
E1	10.5 ± 3.3^c	6.4 ± 1.4^ab	62.1 ± 0.8^a	33.7 ± 2.4^b	26.0 ± 1.3^b
E2	16.4 ± 2.8^d	12.4 ± 2.7^c	66.1 ± 2.5^a	36.2 ± 0.8^b	34.4 ± 2.9^c
Identical letters in the same column indicate no significant difference at p > 0.05 (Fisher's least significant differences).

Moisture and Total Fat Contents

Moisture and fat analyses (Table 4) were conducted separately for the crust and the piece of chicken in the nuggets. The samples prepared with the commercial batter presented greater crust moisture than those with the egg layer. Core moisture values (around 56% g/100 g db) are in agreement with the sensory perception of juiciness, which was found to be greater for samples B1 and B2.

The fat content was higher in the double-coated crusts, although this difference was only significant in the samples prepared with egg. In the chicken pieces, no significant differences in fat content were encountered (all the values were around 3.6 g/100 g db). The lubricity conveyed by the higher fat content of the double-coated samples could also contribute to the sensory perception of juiciness.

Correlation Between Instrumental and Sensory Analysis

Despite the fact that the differences between the samples were small PLS analysis of all the sensory and instrumental data was conducted to examine their correlations. As expected PLS analysis revealed that sensory descriptors could not be well predicted by instrumental measurements. None of the dimensions proved to be significant (<0.001), and the variability explained was low (<50%). The instrumental data were not good predictors of sensory behavior, probably because of the difficulty in parameter evaluation, such as crunchiness by instrumental methods, which only register some aspects of the sensations that a food can produce in the mouth. Although many advances have been made in this field in recent years and the equipment is steadily being perfected, for instance through simultaneous acoustic recording, there are still highly complex sensations that can only be interpreted by a human being. Also, as mentioned above, sensory differences between samples were small, and this could have led to poor correlations for this lot of samples.

DISCUSSION

Despite the proven difficulty of measuring the quality factors of certain foods by solely instrumental means, few studies have employed sensory analysis in their characterization.[4,8,21] The present study chose a food with a complex texture (sandwich-like structure product with several layers, which have very different sensory characteristics) and a final cooking method (microwaving) that did not help to improve the scenario. It also studied two processing factors that a priori would present quite subtle differences: egg or a commercial batter layer before breading, and one or two coatings.

No clear correlation was encountered between the instrumental texture analysis and the evaluation of texture by a descriptive panel, demonstrating the importance of sensory analysis for characterizing this type of complex texture. It was also found that other instrumental measurements, such as fat and moisture content, correlated with the sensory parameters of perception of juiciness and hardness, inviting a revaluation of the employment of oil absorption barrier and moisture retaining substances when designing this type of product, in view of their direct link with its sensory quality.

Despite the constant publication of studies that investigate the possibilities of new ingredients for improving the quality of breaded or battered products, many of which also aim to reduce the negative characteristics that arise when the product is microwaved, these ingredients do not enter mainstream commercial practice, perhaps because the right instruments are not being used to evaluate their effects. Lately, products that can be fried or heated in a conventional or microwave oven for their final cooking stage have come on the market. However, the texture of microwaved product is usually impaired relative to those heated by the other methods. As a result, this option is not adopted by consumers and is still less frequent in the food service sector.

CONCLUSION

The commercial batter (instead of egg) and double coating were both used with the aim of achieving a product with greater structural reinforcement of the external covering. The reasoning was that after the initial pre-frying, this structure would help to keep the crust crunchy even when the final cooking took place in a microwave oven. However, the results indicate that the effects of both were not sufficient to achieve this aim, as they only brought about minimal improvements. The resulting products were still far from possessing the sensory properties of the conventional fried product. The present study highlights the action to improve the quality of certain types of food needs to be taken principally through sensory techniques.

ACKNOWLEDGMENTS

The authors are indebted to the Comisión Interministerial de Ciencia y Tecnología for financial support (Project AGL 2006-11653-C02-01) and to the Ministerio de Educación y Ciencia (Spain) for the grant awarded to the author Á. Albert. They also wish to thank Mary Georgina Hardinge for assistance with the English text.