Imitation Cheese Manufacture Using Rapid Visco-Analyzer and Its Optimization

Abstract

Imitation cheeses in a laboratory-scale (100-g) made by a rapid visco analyzer (RVA) at different stirring speeds (200, 300, 450 rpm) were investigated with a control (Stephan cooker at 1,500 rpm) at 85 °C through functional properties, microstructure and sensory evaluation. Compared to the cheeses made by the Stephan cooker, the cheese made by RVA was significantly (P < 0.05) more green, more yellow and noticeably less white than all other cheeses, which had a decrease in the L* values and an increase in a* and b* values. TPA values of imitation cheese made by the RVA at 450 rpm were lower than that cheese made by other speed. However, the imitation cheeses of 450 rpm was significantly (P < 0.05) higher, which was similar with the fat leakage of imitation cheese manufactured by Stephan cooker. The melt of imitation cheese significantly (P < 0.05) reduced by increasing the stirring speed of RVA. While significant, the melt of the 1,500 rpm cheese was smaller than all other cheeses. Increasing the stirring speed of RVA showed a more effective fat emulsification and a uniform protein matrix. Sensory analysis showed the imitation cheese made by RVA at 450 rpm was similar to the control made by Stephan cooker. Subsequently, the stirring speed, mixing time and heating temperature were optimized at 450 rpm, 7.50 min and 86 °C by response surface methodology (RSM). Results indicated that RVA could make imitation cheeses as well as the functional properties of the control in a laboratory-scale.

Keywords:

• Imitation cheeses
• Rapid visco analyzer (RVA)
• Functional properties
• Microstructure
• Sensory evaluation

Introduction

Imitation cheeses are manufactured by blending various edible ingredients and water to form cheese-like products with the aid of heat and mechanical shear. A sharp increase in the demand of mozzarella cheese as a primary ingredient on pizza has not only caused difficulties in supply, but also increased the cost of the product.Thus, imitation cheeses are widely used because of their cost-effectiveness, the simplicity of their manufacture, and the replacement of selected milk by non-dairy fats or proteins.^[1–3] In order to improve the quality of imitation cheeses, researchers have highlighted the processing conditions, such as the cooker types, shear rate, cooking time, and temperature, which could influence the functional properties of cheese products.^[2,4,5] However, the numerous cooker types and processing conditions make the processors difficult to select an appropriate processing profile for their ingredients blending in the actual production. Taking into account the cost, there has recently been an increased interest in small-scale cheese making, compared with a pilot-scale (4.5 or 2 kg) in a cheese cooker.^[6] Therefore, we may make and analyze imitation cheeses in 100-g batches using a rapid visco analyzer (RVA) in the laboratory to address the problems.

A RVA is the most effective instrument available today for determining the cooked viscous properties of starch, grain, flour, and other foods using its own tailor-made profiles of mixing, measuring, heating, and cooling. Kapoor et al.^[7] showed that a method for making processed cheese in 25-g batches could be developed by the RVA through variable conditions of shear and temperature. The melt and flow properties test for pasteurized processed cheese and pasteurized processed cheese food made by RVA and a twin-screw cooker indicated that RVA melt analysis was correlated with the results from other melt tests and was capable of measuring the properties quantified by other methods.^[8–10] It was also uncovered that the RVA manufacturing time (short versus long) did not have a significant effect on TPA-hardness values for pasteurized processed cheese and pasteurized processed cheese food. Consequently, the modifications used in the RVA method, such as temperature, mixing speed, and time of manufacture, were successful in producing process cheese with functional properties similar to the processed cheese produced in a pilot scale.^[7] However, there are few researches about the effects of RVA on the functional properties of imitation cheeses in a small-scale model.

The aim of the present study was to investigate and compare the texture and microstructure of imitation cheeses made by RVA and a Stephan cooker at different stirring speeds. This study may develop a method for making imitation cheeses using a RVA, increase our understanding of the effects of the equipment on the properties of imitation cheese, and even further, provide more information to the cheese processors.

Materials and methods

Materials

Rennet casein was supplied by Hualing Gansu Dairy Products Co. Ltd., China. The rennet casein contained about 88.57% protein (dry basis) and 11.25% moisture which were determined according to the method of AOAC (2000)^[11]. Soya oil was obtained from China Oil and Foodstuffs Co. (COFCO, China). The other ingredients including trisodium citrate (TSC), disodium phosphate, sodium chloride, and citric acid were of food grade and supplied by the Key Laboratory of Dairy Science, Ministry of Education, China. All the other chemicals were of analytical grade and purchased from Sigma Chemical Co. (St. Louis, MO, USA).

Manufacture of Imitation Cheeses

Imitation cheeses were made with the following (W/W): water, 48.8 g/100 g raw materials, rennet casein, 22 g/100 g raw materials, soya oil, 25 g/100 g raw materials, TSC, 1.9 g/100 g raw materials, disodium phosphate, 0.4 g/100 g raw materials, sodium chloride, 1.2 g/100 g raw materials and citric acid, 0.7g/100 g raw materials. All ingredients were mixed in a twin-screw cooker (model CC-010, Blentech Corporation, CA) at 50 rpm and hydrated at room temperature for 30 min to get a homogeneous paste. A 100-g portion of imitation samples in an aluminum canister paddle was manufactured through a RVA (model RVA-3; Newport Scientific Pty Ltd., Australia). The mixtures were subjected to a temperature of 85°C over 4 min and then were kept at a stirring speed of 200, 300, or 450 rpm for 7 min, respectively. Then the imitation cheeses were transferred to sealed cylinders (25 × 30 mm, diameter × height). The cylinders were cooled to 4°C and stored at 4°C until use. The control cheese was made using Stephan cooker (Stephan Machinery Pty Ltd, Germany) according to Noronha et al.^[2] The temperature was increased to 85°C in 5 min and held for another 5 min, with the wave-cut blade speed of 1500 rpm, scraping paddle speed of 75 rpm.

Composition and Color of Imitation Cheese

The moisture content of imitation cheese was determined in triplicate using an oven-drying method.^[8,12] The protein content was measured by the macro-Kjeldahl method,^[13] and fat content was determined by Gerber method.^[14] The pH was measured by the electrodes of a pH meter (model 9450, Unicam Ltd., CA, England) directly into a small block of imitation cheese, equilibrated at 22°C. All compositional analyses were performed in triplicate.

The color of the imitation cheese was determined using a color measurement instrument (model ZE-6000, Nippon Denshoku Industries Co., Ltd., Japan) fitted with a measuring aperture of 8 mm diameter and a D65 illuminant according to the method of Noronha et al.,^[2] and the L* (lightness), a* (redness), and b* (yellowness) values were recorded.

Instrumental Texture Profile Analysis (TPA)

Textural properties of the imitation cheese were measured using the TAXT texture analyzer (Stable Micro System, Ltd., UK). A P/5-Delrin cylinder probe was attached to the moving crosshead. Imitation cheese samples were cut into 20 mm high and 18 mm diameter by the cheese borer. The textural parameters were calculated according to the method of Sołowiej et al.^[15] All the samples were tempered at 25°C for 30 min before testing. Each sample was analyzed with a pretest speed of 2.0 mm s^–1, test speed of 1.0 mm s^–1, post speed of 2 mm s^–1, two-bite time interval of 5.00 s, and the peak force was recorded at 80% compression using the TPA function of the TAXT software.

Melt Test

The meltability was assessed according to the modified method of Prow and Metzger.^[8] In the RVA melt test, a 25-g disc of imitation cheese was packed into a RVA canister. The temperature was raised from 25 to 90°C in 5 min, and held for 3 min at 90°C, then cooled from 90 to 25°C for 6 min. In the first 2 min, the stirring speed was increased from 0 to 300 rpm, and then was held at 300 rpm until the end of the test. During the holding period, the minimum apparent viscosity was determined.^[7] The apparent viscosity is a measure of how well a processed cheese flows at a fixed temperature, which also has a significant negative correction with flowability of imitation cheese.

Free-Oil Release Test

The free-oil release of imitation cheese was measured in triplicate using the modification filter paper method of Wang and Sun.^[16] The imitation cheese disks with 17 mm diameter and 5 mm thickness were placed on Whatman No. 1 filter papers (Whatman International, Kent, UK) and heated in an atmospheric oven at 110°C for 10 min. Areas of the original disk and oil ring were determined by planimetry, and the percentage area ratio was calculated as the fat leakage.

Microstructure Analysis

Scanning electron microscopy (SEM; Scanning Electron Microscope S-3400N, Hitachi High-Tech Ltd., Japan) was used to provide microstructure information of imitation cheeses which could be related to the functional properties. The method was according to Sołowiej^[15] and Liu.^[17] Small cubes of imitation cheese approximately 1 × 3 × 5 mm cut with a scalpel were fixed with 2.5% (v/v) glutaraldehyde in water for 1 h and rinsed three times with phosphate buffer. After that, all the samples were dehydrated in a graded ethanol series (50, 70, 90, and 100% [v/v]; 20 min per step) and placed in 100% (v/v) ethanol for a whole night. Specimens were critical point dried through CO₂, and then spatter-coated with gold (3 mA, 2 min) in the cryo-chamber prior to analysis in the microscope chamber, at an acceleration voltage of 5.0 kV.

Sensory Evaluation

A sensory evaluation of the imitation cheese was conducted by a trained 10 member sensory panel after cooking. All panelists (employees of the Key Laboratory of Dairy Science, Northeast Agricultural University, Ministry of Education) were seated in separate booths and samples were presented under a red/green light to avoid visual bias. A 10-point hedonic scale (10 = very good; 1 = very poor) was used for shredability, firmness, color, spreadability, stretchablity, and overall impression. After the sensory evaluation, the means were calculated.

Statistical Analysis

At least three separate batches of each imitation cheese were manufactured by a block randomized design in the experiment. Response surface methodology (RSM)^[18] and central composite rotatable design (CCRD) were used to design the experimental combinations for RVA manufacture standardization. Significant differences (p < 0.05) between means were identified using Turkey procedures, unless stated differently. All the statistical analysis were using the General Linear Model procedure of Statistix 8.1 software package (Analytical Software, St. Paul, MN). The model used for RSM as following:

RESULTS AND DISCUSSION

Effect of RVA on Physical Properties of Imitation Cheese

Composition and color of imitation cheese

The mean and SD (p > 0.05) of the moisture, protein, fat, and pH of all imitation cheeses were shown in Table 1. There were no significant differences (p > 0.05) in the overall mean compositional values among the imitation cheeses made by RVA at different stirring speeds and a Stephen cooker. Table 2 shows the color values of the imitation cheeses. The negative values of a* indicated a tendency toward green and the positive values of b* indicated a tendency toward yellow. The color values of imitation cheeses (except a*) made by the different speed RVA were no significant differences (p > 0.05) between each other. Compared to the cheeses made by RVA, the cheese made by the Stephan cooker at 1500 rpm was significantly (p < 0.05) less green, less yellow, and noticeably more white than all other cheeses, which had an increase in the L* values and a decrease in a* and b* values.

TABLE 1 Composition of imitation cheeses manufactured using RVA at a stirring speed of 200, 300, and 450 rpm or Stephan cooker of 1500 rpm

	Stirring speed (rpm)	Moisture (%)	Protein (%)	Fat (%)	pH
RVA	200	48.21^a ± 0.07	19.46^a ± 0.22	24.74^a ± 0.04	5.94^a ± 0.01
	300	48.21^a ± 0.08	19.65^a ± 0.08	24.67^a ± 0.05	5.91^a ± 0.01
	450	48.16^a ± 0.14	19.42^a ± 0.12	24.69^a ± 0.02	5.94^a ± 0.02
Stephan cooker	1,500	48.25^a ± 0.04	19.40^a ± 0.09	24.65^a ± 0.04	5.96^a ± 0.01

^1aMeans with the same superscript letter are not significantly (p > 0.05) different within rows;

²The data (mean ± standard deviation) related to three replicates as described in materials and methods.

TABLE 2 Color values of imitation cheeses manufactured using RVA at a stirring speed of 200, 300, and 450 rpm or Stephan cooker of 1500 rpm

	Stirring speed (rpm)	L*	a*	b*
RVA	200	85.10^b ± 0.39	-1.60^b ± 0.05	13.54^a ± 0.06
	300	85.19^b ± 0.43	-1.62^b ± 0.04	13.52^a ± 0.03
	450	85.47^b ± 0.42	-1.57^a ± 0.05	13.49^a ± 0.05
Stephan cooker	1,500	91.31^a ± 0.48	-1.63^b ± 0.04	8.78^b ± 0.05

^1a–bMeans in the same column with different superscript letters differ significantly (p < 0.05);

²The data (mean ± standard deviation) related to three replicates as described in materials and methods.

In this study of chemical composition of the imitation cheeses (Table 1), there was not a significant difference (p > 0.05) in the weight loss between RVA and the Stephan cooker. The functional properties of the final imitation cheeses could be controlled not only by ingredients, but also by the method of manufacture.^[4,8,19,20] A similar report by Mert et al.^[21] showed the color variation among the imitation cheeses may be related to their microstructure. Increasing the stirring speed indicated a more effective fat emulsification,^[22] and the fat globules embedded in the protein matrix (Fig. 3) making the cheese whiter and less yellow (Table 2).

FIGURE 1 Scanning electron microscope (SEM) images of the imitation cheese, all magnitude at 1000x. The imitation cheeses manufactured using RVA at a stirring speed of 200 (a), 300 (b), and 450 rpm (c) or Stephan cooker of 1500 rpm (d). The white letters F: fat globule; P: protein matrix; C: cavity.

FIGURE 2 Apparent viscosity of imitation cheeses manufactured using RVA at a stirring speed of 200 (RVA-1), 300 (RVA-2), and 450 rpm (RVA-3) or Stephan cooker of 1500 rpm (at a stirring speed of 300 rpm and the temperature of 90°C in the RVA melt test). Means with different letters (a–d) are significantly (p < 0.05) different.

FIGURE 3 Fat leakage of imitation cheeses manufactured using RVA at a stirring speed of 200 (RVA-1), 300 (RVA-2), and 450 rpm (RVA-3) or Stephan cooker of 1500 rpm. Means with different letters (a–b) are significantly (p < 0.05) different.

Instrumental TPA

The variation in textural characteristics of the imitation cheeses manufactured at speeds of 200, 300, and 450 rpm by the RVA and 1500 rpm by the Stephan cooker were shown in Table 3. When the stirring speed of the RVA was increased from 200 to 450 rpm, the hardness of imitation cheese was significantly increased by 4.72%, the adhesiveness was sharply decreased by 20.15%, and the springiness did not have a significant difference (p > 0.05). Compared with the 1500 rpm cheese, the hardness and cohesiveness of imitation cheese made by the RVA at 450 rpm were decreased by 5.75 and 5.97%, respectively, while the springiness did not have a significant difference (p > 0.05).

TABLE 3 Texture profile analysis of imitation cheeses manufactured using RVA at a stirring speed of 200, 300, and 450 rpm or Stephan cooker of 1500 rpm

	Stirring speed (rpm)	Hardness (g)	Adhesiveness (g.sec)	Springiness	Cohesiveness
RVA	200	568.72^c ± 11.80	-486.34^c ± 5.72	0.92^b ± 0.01	0.56^c ± 0.01
	300	570.72^c ± 4.30	-405.41^b ± 13.63	0.92^b ± 0.02	0.58^c ± 0.01
	450	595.55^b ± 4.10	-388.34^a,b ± 7.39	0.93^a,b ± 0.01	0.63^b ± 0.01
Stephan cooker	1,500	631.91^a ± 4.28	-377.56^a ± 6.84	0.95^a ± 0.01	0.67^a ± 0.02

^1a–cMeans in the same column with different superscript letters differ significantly (p < 0.05);

²The data (mean ± standard deviation) related to three replicates as described in materials and methods.

Melt Test

The mean meltability values of imitation cheeses were significantly (Fig. 1), which were described by the minimum apparent viscosity during the holding period in the RVA melt test. Within the range of speeds in the present study of imitation cheeses made by RVA, the relationship between the minimum apparent viscosity (y, cP) and speeds (x, rpm) was best described by a linear model y = 0.9x + 685.3 (R² = 0.968), the melt of imitation cheese significantly (p < 0.05) reduced by increasing the stirring speed of RVA. While significant, the melt of the 1500 rpm cheese was smaller than all other cheeses.

The meltability was one of the important functional characteristics of imitation cheese. Increasing the stirring speed reduced the melt of imitation cheese significantly, it was shown by apparent viscosity in Fig. 1. Ma^[23] noted the meltability of cheeses was related with fat globes. During the processing of imitation cheeses, the stirring and heating resulted in rupture of Van der Waals interactions, and the melted fat formed particles of variable diameter, which were emulsified by rennet casein.^[24] As the examination of the imitation cheeses’ microstructure, increasing the stirring speed reduced the fat globule size and formed a tighter protein matrix, which required more energy to melt, and that was similar to the report by Noronha et al.^[25]

Free-Oil Release Test

Figure 2 showed the fat leakage composition of RVA imitation cheeses and Stephan cooker cheese. As the stirring speeds in RVA method increased, there were no significant differences (p > 0.05) between the imitation cheeses of 200 and 300 rpm; however, the imitation cheeses of 450 rpm was significantly (p < 0.05) higher, which was similar with the fat leakage of imitation cheese manufactured at the speed of 1500 rpm by the Stephan cooker. Free-oil release is a major quality property of Mozzarella cheese; however, there have been few published researches on the factors that govern formation of free-oil.^[26] As shown by the microstructure images, the results are in agreement with the delineation of Everett and Olson,^[27] who verified the oil that was released might originate from some of the larger fat globules which were not stabilized as emulsion droplets in imitation cheeses. Thus, the original disk of low-speed cheese was big, and the percentage area ratio was smaller than the high-speed cheese.

Microstructure Analysis

The microstructure of imitation cheese significantly affected its end-product characteristics, and cheese microstructure was routinely examined using SEM as this technique allowed high-resolution imaging of components.^[28] SEM images of imitation cheeses manufactured using RVA and the Stephan cooker (Fig. 3) showed three distinct structures: spherical fat globules (F), protein matrix (P), and cavities (C). The open spaces represented the fat globules which were extracted during sample preparation and the cavities were formed because of the eliminate-out of water or air.^[17,19]

Figure 3a showed a “honeycomb” structure as the numerous fat globules and cavities were distributed in relatively and even aggregately within the protein matrix at the speed of 200 rpm by RVA. With increasing the speed to 300 rpm, the number of fat globules and cavities seemed to decrease. Figure 3b, which was in agreement with Noronha et al.,^[2] found that increasing mixing speeds led to a decrease in fat globule size and improved their distribution throughout the protein matrix. A uniform protein matrix structure was observed in Fig. 3c and 3d. In addition, the fat globules were smaller and uniform in size, which seemed to embed in the protein matrix, representing a better emulsifying ability.^[3]

Although chemical composition was similar in all imitation cheese products, the variation of TPA (Table 3) was obviously because of its protein matrix which can be seen clearly in Fig. 3. Compared with the imitation cheese made by the Stephan cooker at 1500 rpm, increasing the stirring speed of RVA could reduce the fat globule size and form a uniform protein matrix. Faccia et al.^[30] showed that smaller fat globules might contribute to a firmer protein structure. When the stirring speed increased from 200 to 300 rpm by RVA, the diameter of fat globules became smaller. As the stirring speed increased to 450 rpm by RVA, the caseins adsorbed onto the fat globule, causing an increase in protein–protein interactions,^[31] resulting in the similar hardness values with the Stephan cooker because of the greater resistance to deformation. Moreover, the enhanced protein–protein interactions could strengthen the internal bonds of the imitation cheese, which was described by the all values.^[3,18,32] The larger fat globules conferred adhesiveness to the imitation cheeses, but decreased their springiness, because of the fat–protein interactions.^[33] It was reasonable to state that increasing the agitation speed affected the microstructure of imitation cheese by decreasing the fat globule size and improving their distribution throughout the protein matrix.^[28,34]

Sensory Evaluation

The sensory evaluation of the imitation cheeses was reported in Fig. 4. As shown, the imitation cheeses made by RVA at 450 rpm and the Stephan cooker at 1500 rpm were similar in the parameters considered, showing a good overall impression, color, stretchability, shredability, and firmness. However, the imitation cheeses manufactured at the speed of 200 and 300 rpm by RVA were well in spreadability. Based on the sensory evaluation results, increasing the speed to 450 rpm of RVA and 1500 rpm of the Stephan cooker had no significant difference on the sensory properties of imitation cheeses. The imitation cheeses made by RVA at 450 rpm received similar scores with the Stephan cooker in the sensory evaluation, showing that they were well accepted by the panelists. By sensory evaluation, we demonstrated that manufacture of the imitation cheese by RVA could be better than the Stephan cooker. For the selection of the optimum conditions, the model was analyzed with the given set of process parameters to get the value of over impression. Sensory analysis of the imitation cheese prepared at the optimum condition showed a comparable result in over impression with control.

FIGURE 4 Sensory evaluation of the imitation cheeses considered in the study. The imitation cheeses manufactured using RVA at a stirring speed of 200 (RVA-1), 300 (RVA-2), and 450 rpm (RVA-3) or Stephan cooker of 1500 rpm.

RVA Imitation Cheese Manufacture Standardization

In preliminary studies, a laboratory-scale (100-g) manufacturing model of imitation cheese could be developed by a RVA at around 450 rpm. In order to establish a method for imitation cheese manufacture in the RVA, a CCRD was used.

The overall impression values, under different conditions, are given in Table 4. To optimize the process parameters, a response surface model for the overall impression values of the finished cheeses was developed employing multiple regression techniques

TABLE 4 Responses obtained for different process parameter combinations

Coded variables				Uncoded variables
X1	X2	X3		Heating temperature (℃)	Stirring speed (rpm)	Mixing time (min)		Over impression values
–1	–1	–1		83	435	6.5		7.40
1	–1	–1		87	435	6.5		6.99
–1	1	–1		83	465	6.5		8.38
1	1	–1		87	465	6.5		7.92
–1	–1	1		83	435	7.5		7.46
1	–1	1		87	435	7.5		8.40
–1	1	1		83	465	7.5		7.46
1	1	1		87	465	7.5		8.38
–1.682	0	0		82	450	7.0		6.06
1.682	0	0		88	450	7.0		6.99
0	–1.682	0		85	425	7.0		7.46
0	1.682	0		85	475	7.0		8.87
0	0	–1.682		85	450	7.2		8.40
0	0	1.682		85	450	7.8		9.33
0	0	0		85	450	7.0		9.32
0	0	0		85	450	7.0		8.87
0	0	0		85	450	7.0		8.86
0	0	0		85	450	7.0		8.86
0	0	0		85	450	7.0		9.33
0	0	0		85	450	7.0		9.28

The analysis of variance (ANOVA) showed that the model satisfactorily explained the effect of process parameters on overall impression with no significant lack-of-fit, and the R² was 0.9464, indicating that 94.64% of the variability can be explained by the model. Factor contribution rates obtained from the F-test were in the following order: stirring speed > mixing time > heating temperature.

It was indicated by analysis on the regression model with RSM that there was optimal overall impression value of 9.335 when the stirring speed was 451.46 rpm, the mixing time was 7.50 min, and the heating temperature was 85.57°C. Therefore, 450 rpm, 7.50 min, and 86°C could be chosen to be the optimum parameters in RVA imitation cheese manufacture, mainly because it was easily controlled in practical applications.

CONCLUSIONS

These results suggested that increasing the stirring speed of RVA had a significant influence on the microstructure of imitation cheeses, making it possible to achieve similar functional properties in texture, meltability, free-oil, and sensory evaluation with the Stephan cooker control. The optimum conditions of process parameters for the RVA method were 450 rpm stirring speed, 7.50 min mixing time, and 86°C heating temperature. The data of this study suggested that the imitation cheeses could be made by RVA, which would be of great interest to the researchers because of the small-scale. Further studies are needed to analyze the data obtained during RVA manufacture of imitation cheeses could be used to predict the changes of final products.

Acknowledgments

The authors would like to thank Mr. Li Xiao-Dong and Dai Xian-Qi for their technical assistance, Shanghai Sangon Biotech Ltd., Merck China, and Heilongjiang Academy of Agricultural Sciences.

FUNDING

This work was supported by the project Heilongjiang Province Educational Department General Project and Heilongjiang Province Colleges and Universities in the Industrialization of Scientific and Technological Early Stage of Research and Breeding Project.