3.2. Regression Model of DPCD-Induced Shrimp Surimi Gel
Based on the Box-Behnken design and factor level coding table (
Table 1), a test scheme for the gelation of shrimp surimi induced by DPCD was established (
Table 3). Fifteen test numbers were randomized, and then the gel strength of DPCD-induced shrimp surimi gel was measured using a texture analyzer (
Table 3). The regression model established by response surface regression analysis on the data in
Table 3 was as follows:
Table 4 shows that the predicted
R2 of 0.85 is in reasonable agreement with the Adjusted
R2 of 0.97; i.e., the difference is less than 0.2. Adeq precision measures the signal-to-noise ratio. A ratio greater than 4 is desirable. The ratio of 26.46 indicates an adequate signal. This model can be used to navigate the design space.
The coefficient of regression model (5) was tested for significance, and the results are shown in
Table 5. As can be seen from
Table 5, the effect of
x32 terms (i.e., quadratic term of treatment time) on the strength of shrimp surimi gels in the regression model was insignificant (
p > 0.05). Therefore, the regression model needs to be further optimized by excluding the insignificant term in (5). After removing
x32 from regression model (5) and then conducting response surface regression analysis on the data in
Table 2, the regression model was established as follows:
The coefficient of regression model (6) was tested for significance, and the results are shown in
Table 5. As can be seen from
Table 5, each item in the regression model has a significant impact on the gel strength of shrimp surimi (
p < 0.05). Therefore, each item in
Table 5 can be retained in the regression model (6).
Regression model (6) was used to predict the gel strength of shrimp surimi in the 15 tests in
Table 2, and linear correlation analysis was conducted between the predicted value and the experimental value of shrimp surimi gel strength. The results are shown in
Figure 5. It can be seen from
Figure 5 that the predicted value of gel strength presented a suitable linear correlation with the experimental value, with determination coefficient
R2 = 0.99 and root mean square error RMSE = 3.72. This indicates that regression model (6) could explain 99% of the changes in the gel strength caused by the mixing pressure, temperature, and time, and the regression model (6) had a high reference value. The analysis of variance and the lack of fit test of the regression model (6) were performed. As can be seen from
Table 6, the regression model was significant (
p < 0.05), while the loss of fit was not significant (
p > 0.05), which indicates that the fitting of regression model (6) was feasible within the variation range of independent variables (−1, 1).
The results of the coefficient significance test, coefficient of determination (
R2), analysis of variance, and loss-of-fit test of the regression model (
Table 6) fully indicated that regression model (6) could be used to predict the gel strength of shrimp surimi under different mixed treatment pressure, temperature, and time and optimize the process parameters.
3.3. Change of Surimi Gel Strength during DPCD Treatment
Since the Box-Behnken test design encodes the pressure, temperature, and time factors for the DPCD treatment, eliminating the influence of each factor unit, they were all varied in the range [−1, 1]. Therefore, the gel strength of each shrimp surimi was calculated sequentially based on the absolute values of coefficients in regression model (6). According to the absolute values of each coefficient in
Table 7, the order of magnitudes affected by the strength of each shrimp surimi gel was pressure once term (
x1), time once term (
x3), temperature once term (
x2), pressure and temperature cross term (
x1x2), pressure squared term (
x12), temperature and time cross term (
x2x3), temperature squared term (
x22), and the intersection of pressure and time (
x1x3). Where the pressure once term (
x1), temperature once term (
x2), and time once term (
x3) were positively related to the gel strength of shrimp surimi, while the pressure to temperature cross term (
x1x2), pressure squared term (
x2), temperature to time cross term (
x2x3), temperature squared term (
x3), pressure to time cross term (
x1x3) were negatively related to the gel strength of shrimp surimi.
Based on the regression model (6), the interaction among DPCD pressure, temperature, and time was analyzed (
Figure 6). In the interaction figure, if any two factors are parallel to the two curves at the vertical and horizontal intersection, it indicates that there is no interaction between the two factors; if two curves are intersected, it indicates that there is an interaction between the two factors [
22]. It can be seen from
Figure 6 that both curves were crossed between any two factors, indicating that the interaction between pressure and temperature, pressure and time, and temperature and time of DPCD treatment on the strength of shrimp surimi gel were significant (
p < 0.05). Therefore, the effect of DPCD treatment on gel strength was not caused by a single factor but by the combined action of temperature, pressure, and time.
In order to investigate the influencing of three factors (pressure
x1, temperature
x2, and time
x3) on the strength of shrimp surimi gels in regression model (6), one of them was fixed as intermediate level (i.e., the zero level of coded values), three-dimensional surface plots were made by using the surface characterizer of JMP10.0 software, and then orthographic projection was made (
Figure 7).
The induction of protein gel formation by DPCD is the result of the combined action of the two effects. One is the thermal effect; under the induction of heat, myosin will gradually denaturate, stretch, and aggregate to form gel [
23]. The other one is the molecular effect of CO
2; at a high-pressure state, CO
2 dissolves in water to form carbon acid, and carbon acid dissociates hydrogen ions, reducing the system’s pH. Hydrogen ions also make the myosin molecules positively charged. At the same time, CO
2 is a non-polar molecule, which can interact with the hydrophobic group of proteins, and denaturate myosin through electrostatic interaction, hydrogen bond, hydrophobic interaction. Myosin molecules, in turn, cross-link each other through hydrophobic interactions, disulfide bonds, nondisulfide covalent bonds to form a three-dimensional meshwork gel [
16,
17].
As can be seen from
Figure 7B, the gel strength of shrimp surimi increased rapidly under the condition of low temperature with the increase in pressure, while the gel strength of shrimp surimi slowly increased with increasing pressure at high temperature. This may be because, at low temperatures, CO
2 molecules moved slowly, and increasing the pressure increased CO
2 density, and the number of CO
2 molecules that interact with the surimi increased. At the same time, the pressure increase also made CO
2 more easily permeated and diffused into the surimi, and the effect of CO
2 molecules was enhanced so that the gel strength increased rapidly [
24].
It can be seen from
Figure 7C that, under high pressure, the gel strength of shrimp surimi increased slowly with the increase in treatment temperature, while under low pressure, the gel strength of shrimp surimi increased rapidly with the increase in treatment temperature. The molecular effect of temperature on CO
2 has two sides [
16,
17]. On the one hand, the increased temperature would aggravate the thermal movement of CO
2 molecules (referred to as the thermal motion effect), which would benefit its penetration and diffusion into the surimi, and make sufficient interaction with myosin. On the other hand, the increasing temperature decreased CO
2 density (referred to as the density drop effect), which was unfavorable for CO
2 dissolution in surimi. Under the condition of high pressure, the density of CO
2 was relatively high, and the molecular thermal motion effect caused by increasing temperature was slightly greater than the density drop effect, so the increase in temperature on gel strength was small. Under low-pressure conditions, the CO
2 density was small, and the molecular thermal motion effect caused by increasing temperature was larger than the density drop effect, which would aggravate the molecular thermal motion of CO
2 and favor its penetration and diffusion into surimi, thereby enhancing the gel strength.
It can be seen from
Figure 7E,H that the gel strength of shrimp surimi increased gradually with increasing temperature and pressure regardless of long or short treatment time, which illustrated that both temperature and pressure were important factors affecting the gel strength of shrimp surimi. This may be because increasing the pressure increased the molecular effect of CO
2, promoted myosin denaturation, and increased the strength of the shrimp surimi gel, regardless of the treatment time [
13]. Regardless of treatment time, increasing temperature exacerbated CO
2 molecular thermal movement, accelerated its penetration and diffusion into surimi, and thus enhanced the gel strength.
From
Figure 7F,I, it can be seen that no matter high pressure or low pressure, whether high temperature and low temperature, surimi strength showed an obvious increasing trend with the treatment time, which was because the interaction between CO
2 and myosin molecules was more sufficient with the extension of treatment time, thus enhancing the gel strength. In addition, it was also reported that DPCD treatment was more effective than heat treatment in inactivating cathepsin L activity that causes surimi gel deterioration. At the same time, heat treatment was more effective than DPCD treatment in inactivating TGase activity that enhances gel strength, which also promotes DPCD shrimp surimi gel strength to be stronger than heat-induced shrimp surimi [
25].
3.5. Effect of DPCD on the Quality of Shrimp Surimi
It can be seen from
Table 9 that compared with heat induction, DPCD reduced the loss of moisture, crude protein, and ash in the shrimp surimi gel. This may be because the preparation of shrimp surimi gel by heat induction was made by boiling water, during which part of the water-soluble proteins and minerals may dissolve in water, resulting in the loss of proteins and minerals in the gel. DPCD-induced shrimp surimi gels were carried out in an anhydrous CO
2 environment. Although DPCD is somewhat extractable for moisture, it is not extractable for proteins and minerals, and the process is static, and only small amounts of water-soluble proteins and minerals may be carried out by moisture during unloading [
26]. Whereas there was a small loss of protein from the gel and essentially no loss of minerals. Correspondingly, the weight loss of heat-induced shrimp surimi gel was 323.92% greater than that of DPCD.
Sensory evaluation showed that the sensory scores of DPCD-induced shrimp surimi gels were significantly higher than that of heat-induced (
p < 0.05), which was mainly because heat induction was performed in a water bath, and more odor and taste substances were easily lost. In addition, the slow heat transfer rate and uneven mass transfer induced by heat also lead to the loose structure of the gel [
27]. In contrast, DPCD induction was carried out in an anhydrous environment with less loss of taste substances. Although DPCD was extractable for volatile odorants, it was a static treatment and, therefore, only had a small loss during unloading pressure. The DPCD also denatured the protein sufficiently, which gave the shrimp surimi gelatin a bright pink color. In addition, the surface tension of CO
2 in the supercritical state is zero, which made the dissolution and diffusion of CO
2 in the shrimp surimi more uniform, thus forming a more uniform and dense gel morphology.
Gel strength and water-holding capacity of gels are important indicators for evaluating the quality of gels. As can be seen from
Table 9, the gel strength and gel water-holding capacity of DPCD-induced shrimp surimi were significantly higher (
p < 0.05) than those of heat-induced shrimp surimi gel. This is mainly because heat-resistant alkaline proteases (optimum 50–70 °C) present in shrimp surimi degraded myofibrillar proteins during heat-induced shrimp surimi gel, greatly reducing the elasticity and strength of the gel [
3]. The temperature of DPCD to induce the formation of gels from surimi was below 55 °C; moreover, DPCD has a passivating effect on some high-temperature-resistant alkaline proteases [
28]; Therefore, the gelation deterioration was inhibited. In addition, the gel strength is closely related to the microstructure of the gel. Generally speaking, the smaller the pore size and uniform distribution of the gel network, the bigger the gel strength and water-holding capacity of the gel will be. According to the sensory evaluation and microstructure observation of shrimp surimi gels, the gel network of heat-induced samples had large pore sizes and uneven distribution, while the gel network of DPCD-induced shrimp surimi gels had small pore sizes and uniform and compact distribution. Therefore, the gel strength and water-holding capacity of shrimp surimi induced by DPCD were relatively high.
The TPA properties of the shrimp surimi are shown in
Figure 9B. The shrimp surimi gels were characterized by high chewiness, hardness, adhesiveness, springiness, and small cohesiveness. The chewiness, hardness, adhesiveness, and springiness of the DPCD-induced samples were significantly higher than that of the heat-induced ones, which corresponded to the gel strength results. This may be because DPCD promoted the formation of a fine and uniform gel network structure, which increased the water-holding capacity of the gel, so the texture properties of samples increased.
The low-field nuclear magnetic resonance (LF-NMR) curves (
Figure 10) of shrimp surimi gels had four peaks whose relaxation times were denoted by
T2b1,
T2b2,
T21, and
T22, corresponding to different states of water in the gel, respectively.
T2b1 and
T2b2 (0–10 ms) represent bound water tightly bound to protein molecules.
T21 (30–500 ms) represents immobilized water that entrapment by macromolecules inside the shrimp surimi gel. The
T22 (500–2000 ms) represents free water in the shrimp surimi gel system [
29]. The DPCD shrimp surimi gel lacked one
T22 peak (i.e., free water) than the heat-induced gel, indicating that free water in DPCD was below the detectable limit. The relaxation time
T2 is related to the binding force and degree of freedom of hydrogen protons in the sample. The larger the
T2, the smaller the binding of water molecules and the greater the degrees of freedom [
20].
As can be seen from
Figure 9,
T2b1,
T2b2, and
T21 of the shrimp surimi gel induced by DPCD were significantly shorter than those of the heat-induced shrimp surimi gel (
p < 0.05), indicating that DPCD increased the binding capacity of the shrimp surimi gel to bound water and immobilized water.
A2 represents the relative content of water molecules in different states. It can be seen from
Figure 9 that the immobilized water (
A21) was the most important water in shrimp surimi, accounting for 89.93% and 93.55% of the heat-induced gel and DPCD gel, respectively. The relative contents of the bound water and immobilized water in the DPCD-induced gel were both higher than those in the heat-induced gel. It was further indicated that DPCD reduced the loss of bound water and immobilized water in shrimp surimi gel. This is closely related to the microstructure of shrimp surimi gel. It was known from sensory evaluation and microstructural analysis that heat-induced shrimp surimi gels had loose tissues with a large pore size of gel network, which easily made the transformation of immobilized water into free water for loss, while DPCD-induced shrimp surimi gels had dense tissues with uniform and small pore size of gel network, which locks immobilized water. This result corresponds to those of water-holding capacity, gel strength, and microstructure. It was also well illustrated that the shrimp surimi gel induced by DPCD was significantly superior to that induced by heat in terms of quality.
As shown in
Figure 9A, the
L*,
a*, and
b* values of DPCD-induced shrimp surimi gel are significantly higher than those induced by heat (
p < 0.05), which may be related to the microstructure of shrimp surimi gel formation. The shrimp surimi gel induced by DPCD is more uniform and dense than that induced by heat, which makes it more reflective to light, thus making its
L* value and
a* value correspondingly larger.
The appearance and SEM images of different shrimp surimi gels are shown in
Figure 11. It can be seen from the appearance image that the gel structure of the heat-induced shrimp surimi was relatively loose (
Figure 11A), and the surface of the product had the characteristics of large cracks and rough surfaces. The DPCD-induced shrimp surimi gel had a smooth surface and tight structure (
Figure 11B). From the SEM images, it can be found that the holes in the network of heat-induced shrimp surimi gel were coarse, and the surface was bumpy. At the same time, DPCD gel had a fine network with fine and uniform pores. This further indicated that DPCD promoted the formation of a dense gel network structure and improved the gel quality of shrimp surimi. Rawdkuen, Sai-Ut, Khamsorn, Chaijan, and Benjakul [
30] suggested that regular structure was intrinsic to stronger gel strength. Boiling heating in water usually caused the network pore size of surimi gels to be large and non-uniform due to the slow heating rate and non-uniform mass transfer [
31]. During the treatment of DPCD, the surface tension of CO
2 in the supercritical state was zero, which made its dissolution and diffusion in the sample relatively uniform, and it was beneficial to form a gel network structure with relatively small pore size and uniform and dense. Therefore, DPCD could induce shrimp surimi to form a gel with a better microstructure.