2.1. Composition Analysis
The starch mineral composition can be of decisive importance for its industrial application, because the degree of phosphorylation determines the physicochemical properties of the polymer [
31,
32,
33]. Moreover, it is likely that the level of divalent cations, such as calcium and magnesium, has a significant influence on starch gelatinization properties through the ionic cross-linking of phosphate starch esters [
34,
35,
36]. The investigated starch samples were characterized by a low ash content (
Table 1). The higher values, compared to the SO sample (control; two-year average 0.20%), were recorded for both PS-GH and FT (two-year average by 40% and 65%, respectively). All the obtained values were within the range reported for the PS [
15,
37,
38,
39] and within the limit of 0.5%, recommended for industrial grade A starches [
40].
The non-carbohydrate compound found in relatively high amounts in the PS was phosphorus. Its high content in the PS is desirable in order to obtain a high viscosity paste [
41]. The content of phosphorus in the analyzed PS (
Table 1) ranged from 540.25 mg/kg for SO-II (second year of cultivation) up to 745.26 mg/kg for FT-II. Taking into account the two-year average content of P in PS, it can be concluded that the hydroponic cultivation of plants improves the accumulation of this element in starch. In the soil environment, the phosphorus uptake by plants was limited by the chemical sorption of this element. Chemical sorption of phosphorus takes place mainly with iron, aluminum and manganese compounds. Thus, for any plant, phosphorus uptake from the soil is much more difficult than from the nutrients in hydroponics. Therefore, the phosphorus content of starch from hydroponic potatoes was higher than in soil. We also observed a statistically important difference in the content of this macronutrient in the polymer between different hydroponic systems (closed vs. open one). The phosphorus content in GH (closed system) was on average 6.28% lower than in FT (open system). The presented values are comparable with the values given in other publications [
6,
21,
42].
The PS contained natural metal cations bound by ionic forces to the phosphate ester groups. The level of divalent cations, such as calcium and magnesium, has a significant influence on the pasting properties of starch, possibly by ionically cross-linking the starch phosphate esters [
34,
35,
36,
43]. De Willigen et al. [
44] reported that the characteristics of ionically cross-linked starch paste should be similar to those of covalently cross-linked starch.
The content of calcium in potato starch varied widely, from 21.89 to 98.41 mg/kg (
Table 1). These values are consistent with those reported by other researchers [
16,
36,
38,
45]. The average two-year content of calcium ions in the PS extracted from plants grown in both NFT systems is significantly lower than for reference sample (SO) (for GH by 77%, whereas for FT by 71%). Such a large difference, although statistically significant, was not recorded between GH and FT.
The content of magnesium ions in the studied starches ranges from 46.26 to 73.54 mg/kg (
Table 1), which is consistent with the literature data [
36,
38]. Additionally, for this element, the average two-year content in the PS isolated from the plants grown in the soilless system is significantly lower than for SO (for GH by 34.5%, and by 31.2% for FT). A slight difference of 2.37 mg/kg was observed between the open and closed NFT systems.
The content of the monovalent metal ions (K
+, Na
+) ranged from 524.35 to 920.05 mg/kg and from 6.62 to 25.02 mg/kg, respectively (
Table 1). The data related to the content of these elements in the PS were already published [
34,
36,
38]. Additionally, in the case of the content of the minerals mentioned above, a difference was observed between the PS isolated from the plants grown in soil (SO) and by the hydroponics method. The K
+ content in SO sample was lower on average by approximately 37.5% than in other samples. On other hand, for the Na
+ content, a reverse situation was observed—for the GH sample, it was over two times lower, whereas and for the FT sample, it was one and a half times lower than for the SO. In the case of these ions, the influence of the NFT system (open/closed) on their accumulation in the starch granules was also visible. In the closed system (GH), the content of potassium ions was higher, and in the case of the sodium ions, it was lower than in the open system without nutrient recirculation (FT). According to the literature data [
14,
46,
47], the cations content was relatively higher in the PS with an elevated phosphorus content. The results of this work seem to confirm these observations, because the sum of the content of calcium, magnesium, potassium and sodium in the analyzed starches is 991.15, 979.47 and 756.00 mg/kg for GH, FT and SO, respectively, while a significantly higher P content is noted for the starch isolated from potato tubers from hydroponic cultivation (GH and FT). Sodium is an element commonly found in soil, but this element is not introduced into the nutrients in hydroponic cultivation. Therefore, the sodium content in starch isolated from the potato tubers grown in soil is much higher than that of the starch from the NFT crops. In addition, there is a competition (antagonism) between the potassium ion (K
+) and the sodium ion (Na
+) in plants during root uptake. Since the uptake of potassium from nutrients is much easier than from soil, the K versus Na interaction in soil is weaker than in nutrients. Divalent ions and potassium are antagonistic. The high content and easy uptake of potassium in hydroponics weakens the uptake of Ca
2+ and Mg
2+. These conclusions also confirm the results obtained for the molar fraction of mineral elements (
Table 1).
An important characteristic of starch is the amylose (AM) content. Due to the very low lipid content of potato starches, the observed AM content values are equivalent to both the amount of apparent amylose and the amount of total amylose. In the PS samples tested, the AM content ranged from 26.18% to 27.82% (
Table 1), which is consistent with values reported by other researchers [
14,
15,
33,
38,
48,
49,
50]. The lowest AM content, among the studied samples, was characterized by FT starch (two-year average 26.34 g/100 g). The values higher by 0.70 g/100 g and 1.48 g/100 g were observed for SO and GH samples, respectively, which correspond to an increase in the AM content by 2.66% and 5.62%, compared to FT. The ratio of amylose to amylopectin (AP) ranges from 1:2.59 (GH) to 1:2.80 (FT), and reflects the distribution (spread) of the AP content [
6]. Our results confirm that the ratio of AM to AP is a relatively constant characteristic of the plant botanical variety, but can be modified by cultivation [
14]. The range of amylose content that we observed (SD/mean = 0.027 for all starch samples from the potato variety ‘Vineta’) was as expected.
The low protein (and also ash) content of the investigated starches, ranging from 0.26 g/100 g to 0.33 g/100 g (
Table 1), indicates its purity and is consistent with the data reported by other authors [
6,
48]. The lowest protein content (two-year average nitrogen content) of 0.043g N/100g was recorded for the SO sample. The starch obtained from tubers grown under the NFT system contained statistically more of it (0.046g N/100 g and 0.053 g N/100 g for GH and FT, respectively).
2.2. Starch Color
Color is one of the most important visual quality parameters for both raw materials and products prepared on their basis. The color of the raw material can provide information on the chemical composition of the product, and thus determine its suitability for processing. Moreover, its variation confirms changes that occur in the structure or chemical composition of the raw material or product during processing or storage.
The results of the color measurements of the tested potato starch samples are presented in
Table 2. The color parameter values obtained are similar to those presented in the literature [
51,
52]. Regarding the lightness (L*), it intensifies for all the starches isolated from potato tubers from the NFT crop; in relation to the starch from the SO object, the highest L* value is marked by the GH starch (two-year average L* = 92.05), while the SO starch has an L* value 8.43 lower and the FT starch by 2.69 (
Table 2).
The change in the plant cultivation system from soil to soilless is significantly reflected in the color a* (greenness to redness) and b* (blueness to yellowness) parameters (
Table 2). For the starch isolated from plants from the NFT cultivation, the a* parameter takes negative values, indicating a shift in color towards greenness, whereas for the SO sample, the color shifts towards red (a* > 0). The b* parameter for all the tested starches obtained positive values; however, for the GH and FT starches (b* 2.7), it shifted more towards blue than for the SO starch (b* = 6.06).
From the color parameters L*, a* and b*, the total different color ΔE (
Table 2) was calculated, which reflects the color difference between the two compared samples. The values of ΔE 2.3 corresponds to a just noticeable difference [
53]. When considering the effect of the cultivation method on this parameter, the lowest ΔE values, and therefore the least color difference, were recorded between the hydroponic cultivated starches (
Table 2). The ΔE values obtained for GH in relation to FT (1st year of cultivation) indicated that there was no difference in color between the studied samples that could be noticed by the observer (ΔE = 0.9). There is a noticeable color difference between the other samples and this is most clearly evident in the comparison of the PS from hydroponics to SO. It should also be noted that the year of cultivation is important for this parameter.
The hydroponic cultivation also causes a noticeable decrease in the saturation of the potato starch color, as evidenced by the chroma C* value (
Table 2). Additionally, the whiteness index (WI) confirms that whiter starch was obtained from objects where the soilless cultivation of plants was applied (WI increase for SO ≈ 100%, compared to GH and FT) (
Table 2).
It should be emphasized that all the aspects related to the chromatic attributes of the tested starches confirmed the influence of the type of applied potato plant cultivation method—its change from soil to hydroponic. The starch samples obtained from the tubers grown in soil are darker and grayer with red and yellow tones, compared to the other samples, which can be concluded from the lower values of the L* parameter and higher values of the a*, b*, C* and WI indices.
2.4. Water Binding Capacity (WBC), Welling Power (SP) and Starch Solubility (S)
The temperature dependence of the WBC, SP and S parameters obtained for the analyzed PS samples is presented in
Table 3.
The WBC was influenced by both the temperature and cultivation method. At each measurement temperature, SO had the highest WBC values. The increment of the WBC values between the lowest and the highest measurement temperature was 74.14 and 68.26 g/100 g for GH and FT, respectively, compared to 75.81 g/100 g for SO. The increase in the WBC values with the increasing temperature can be due to gelatinization, which breaks the weak associative bonds in the amorphous region of the starch granules and allows for increased hydration [
58]. The WBC values obtained were slightly lower than those reported in the literature (77.2–89.0%) [
11,
59].
The SP indicates the interaction between the amorphous and crystalline region of the starch granules. At the pasting temperature, the starch granules have limited swelling properties and therefore only a small amount of starch is dissolved, but at a higher temperature there is an increase in the SP value and a large amount of starch leaks out [
60]. The highest increase in the swelling power (7.5 to 8 times) was observed when the temperature was increased from 60 to 70 °C, with the most pronounced increase for the GH sample (
Table 3). At the highest temperature (90 °C), the FT (two-year average 92.70%), GH (two-year average 100.46%) and SO (two-year average 105.15%) starches showed large differences in the SP (
p < 0.05). At 80 °C, the differences in the SP of the tested starch samples was relatively low, but still it was statistically significant. At the other temperatures, the swelling power of the SO was higher than that of the other samples.
By analyzing the obtained results, it was noted that the calcium content was positively correlated with the swelling power at all temperatures, which can be explained by the across-linking of starch chains by covalent bounds created by this divalent ion. A similar observation was made by other authors [
61] (therefore, the calcium content contributes with the water diffusion to the coarse particles and improves the swelling of the starch granules).
Solubility (S) is the measure of the amount of solutes that were washed out of the starch granules when measuring their swelling ability. The starch solubility increased with the increasing temperature. At high temperatures, an increase in the S value indicates an increase in the amount of solute amylopectin, the amount of which increases dramatically when the granules are ruptured [
62]. The solubility of starch was significantly affected by both the temperature and the method of potato plant cultivation (soil/soilless system) (
Table 3). An increase in the solubility with an increasing temperature was observed in every starch sample tested, with the most intense increase between 60 and 70 °C (temperature close to the pasting temperature—
Table 4). SO solubility increased by 5.6 times, while for the PS samples from both types of hydroponic systems the increases were not as pronounced (3.6 and 4.7 times for GH and FT, respectively). Between the temperatures of 70 and 80 °C, the intensity of S increase was not so significant. In the range of these temperatures, the smallest increase was recorded for the SO (about 1.5), and about 2.7 times for the remaining samples. The solubility of SO was 24.14 g/100 g and for the remaining samples these values were lower by 1.4 and 1.9 g/100 g for GH and FT, respectively.
For all the investigated parameters (with few exceptions, namely WBC at 80 °C and S at 60 °C), a statistical difference was recorded between the I and II cultivation year (
Table 3).
2.5. Pasting Properties
The results of the pasting properties of 5% potato starch suspensions in water are presented in
Table 4 and
Figure 1. From observing the pasting curves, it can be concluded that three distinctive courses can be observed for the starches grown in different conditions (GH, FT and SO), with the largest discrepancies observed between the cultivation year for the SO starches. The pasting temperatures (PTs) of all the investigated PS was within a rather narrow range, 69.5–72.1 °C, which are in the range reported in the literature [
63,
64] and also similar for oat starches [
65].
This parameter was affected by the cultivation method, but not by the year of cultivation. The peak viscosity (PV) was varied significantly among the samples, with the lowest values attained by samples grown in soil (SO average 612 BU) and the highest values were attained by the samples grown in the foil tunnel (FT, 1253 BU), clearly indicating the importance of the growing method. Additionally, the differences were spotted for the PVt (time needed to reach PV) and PVT (temperature at PV) parameters, and the lowest values were observed for FT starches, which were characterized by a sharp increase in viscosity followed by a dramatic decrease in it (high values of BD and BD%). Due to the presence of a covalently bound phosphorus moiety [
41], the PS can reach high maximum viscosity values, much higher than for cereal starches [
63,
64,
66], where phosphorus is incorporated in the form of phospholipids restricting granular swelling, and thus limiting maximum viscosity.
In this research, that a well-known correlation between a phosphorus content and starch paste viscosity was observed [
32,
41,
48,
54]. The obtained PV values for SO starches were lower than for other cultivation methods; in fact, their PV values were similar to that observed for cereal starches [
65].
As previously mentioned, the pasting of FT starches was characterized by a rapid growth of viscosity at the first stage of this phenomenon, which was reflected by the appropriate parameters of the logistic model [
67], namely the highest values of V
peak and s, and the lowest r V
peak can be identified with PV, but they were generally slightly lower than the corresponding PV values, with the exception for SO I and II starches. Another parameter applied in this model, r, was the time needed to reach half of the V
peak value, and the highest values were observed for SO starches, indicating a rather slow swelling of the granules, whereas the lowest values were calculated for the rapid viscosity growth observed for the FT samples.
The last, dimensionless parameter of the model (s) is related to granular swelling. As can be observed for slowly developing viscosity SO starches, the value is almost half that of the GH, and over two times lower than for the FT starches.
Starches subjected to the combined action of elevated temperature and shear forces, decreased their viscosity during the holding period, which was manifested by appropriate MV (minimum viscosity), BD (breakdown), BD% and HPSI (hot paste stability index) values. This drop in viscosity was related to the disintegration of swollen starch granules, and can describe the resistance of swollen starch granules against extreme conditions. It was particularly noticeable for FT starches, where a dramatic decrease in viscosity was observed. On the other hand, other starches were characterized by a rather low viscosity drop (small disintegration of swollen starch granules), and by the relatively high stability of viscosity during holding at elevated temperatures (high values of HPSI and low BD and BD%
Table 4). A convenient method to compare this decrease in viscosity for different starches (or for analyses performed using different procedures or equipment) is the BD% value (and at the cooling stage, an increase in the viscosity could be described by means of SB%). The BD% values reported for the potato starches varied in a broad range, from 11.7 to 58.3 [
63,
64], indicating different responses on shearing at elevated temperatures. The calculated BD% values for SO starches (8.8%) in this research were below the lower limit of this range, indicating good starch paste resistance (that was also confirmed by high HPSI values), whereas others (FT and GH) were within it. For the starches of a different botanical origin, the BD% were as follows: waxy maize (50.5–73.1%), maize (12.8–43.7%), wheat (20.3–32.0%), oat (35.9–36.8%), tapioca (71.2%) and pea (59.3) [
65,
67,
68,
69].
In order to describe hot starch paste stability, HPSI can also be applied (
Table 4). The values calculated in this research were found within a broad range (64.6–96.3%), with the highest value observed for SO starches indicating their good stability. The HPSI values for waxy maize starch were 89.0% and 76.4% for regular maize [
69].
At the end of holding, the MV value was observed (from 492 to 819 BU) and then, as a result of cooling (and the creation of hydrogen bonds), an increase in viscosity was observed of up to TV (from 848 to 1375 BU). This increase in viscosity is described as a setback (SB), and could be related to the starch retrogradation. The SB% values obtained in this research varied from 36.4 (SO-I) to 48.1% (FT-I). The calculated SB% (based on literature data) for the potato starches varied from 11.8 to 24.5 [
63,
64].
2.6. Thermal Properties
The functional properties of starch make it one of the most widely used raw materials in the food industry. Its technological potential can be further enhanced by various modifications or the addition of nutrients, which affect the physical and functional properties of a given product. Among the many characteristic properties of starch, one of the most important is the phenomenon of pasting and the subsequent retrogradation of starch. Starch pasting occurs during many food processing operations and has a significant impact on the properties of the resulting products. Starch granules, as a result of water absorption at elevated temperatures, swell and disintegrate, releasing AM, which is poorly soluble in water, and AP. The pasting temperature and the process itself depend mainly on starch type and the amount of available water. The obtained starch pastes and gels are subjected to the retrogradation process. The effect of this phenomenon is the transition of biopolymers from a form of paste into a partially ordered form, and the formation of a crystalline network. During this process, water molecules are expelled from the biopolymer network as a result of the reduction in intermolecular spaces. This leads to structural changes during storage, resulting in the increased turbidity of pastes and gels, increased gel stiffness, syneresis, as well as bread staling.
The results of the thermodynamic analysis of starch gelatinization and retrogradation are presented in
Table 5. Small differences in the gelatinization temperatures were observed. Only in the case of T
Pg and T
Eg did the samples slightly differ, which indicated a different course of the gelatinization process, probably related to a significant difference in the SP values, especially at higher temperatures. The values of all the temperatures analyzed were relatively high and were at or above the upper limits of the values reported in the literature for PS [
33,
63]. The values of gelatinization enthalpy ranged from 15.59 J/g to 17.2 J/g and were not significantly different from a statistical perspective. These values were within those reported in the literature for PS [
33,
63].
As the starch paste cools, AM forms double helixes composed of 40–70 glucose molecules and AP forms crystalline structures [
33,
63]. This process, referred to as starch retrogradation, is associated with the formation of hydrogen bonds between starch chains [
33].
The values (temperatures, enthalpies) obtained with DSC (
Table 5) characterizing this process were generally lower than those characterizing the gelatinization process. In the case studied, the values of the temperature of the onset of transformation (T
Or) ranged from 44.6 °C to 50.13 °C and depended on the method of cultivation. The two samples from hydroponic cultivation (GH and FT) did not differ from each other. The peak temperature (T
Pr) value did not differ, while a similar relationship was observed for the end of the transformation (T
Er) as for the beginning (T
Or), but the variation was slightly smaller. Significant differences were found for the enthalpy of retrogradation.
Samples from conventional cultivation (SO) were not different from the GH sample, while both were significantly different from the FT one. A similar relationship was characterized by the degree of retrogradation (R%). This indicated, on the one hand, the similarity between traditional and hydroponic cultivation, and on the other hand, it indicated the significant effect of the recirculation of the nutrient solution on the PS properties and its behavior during the retrogradation process. The stability of starch and its products strongly depends on their composition and parameters characterizing the storage site (relative humidity and temperature) [
70,
71]. When determining the appropriate storage conditions, the vitreous transition temperature (T
g), a property extremely important from the point of view of the transformations occurring in food, plays a crucial role. It characterizes the change in mobility of the water contained in food products and is related to the interactions between the water molecules and macromolecules of other food components. The values of the temperatures characterizing the phenomenon of vitreous transition and the subsequent melting peak are summarized in
Table 5. In this case, no statistically significant differentiation of samples was found. Only in the case of the onset of vitreous transition and the enthalpy of melting, slightly lower values were found for the conventional cultivation (SO), than for the other two samples derived from hydroponic cultivation (GH and FT).
2.7. Crystallinity
The crystallinity of starch granules can reveal important information about the internal structure and the type of amylose chains distribution within the granules. There are two distinct crystalline polymorphic forms: the A-type (B2 space group) [
72], mostly found in cereal starches, and the B-type (P6
1 space group) [
73,
74], observed in tuber starches (PS is in that group) and high-amylose cereal starches. As starch is composed from mostly amorphous amylose and semicrystalline amylopectin, the powder XRD pattern consists of amorphous part and crystalline peaks (compare Figures 1 and 3 in [
75]). In our case, the XRD patterns of the analyzed starches are presented in
Figure 2a. As can be seen from
Figure 2a, in all the cases, the shape of the curves is similar. The crystalline diffraction peaks can be seen in all the cases, and the best patterns are observed for GH I and GH II starches.
Figure 2b shows the XRD pattern for GH II after data processing (background and amorphic part subtraction). In
Figure 2b, one can recognize crystalline peaks, the most important are 5.74, 10.10–11.74, 14.38, 15.18, 17.26, 19.86, 22.40, 24.22 and 26.36°. Most of them indicate a B-type structure, but a peak at 22.40° could also indicate some fraction of the C-type [
75,
76]. The relative crystallinity of the samples was calculated by comparing the crystalline area with total area of the peaks in the 2θ range 4–30°, using the Equation (2) from the paper of Frost and co-workers [
77]. The results are summarized in
Table 6. As can be seen from the table, the highest values are found for the starches obtained from the potatoes cultivated in a greenhouse (GH I and II), which is consistent with the highest amylose content (compare
Table 1). The other samples have a significantly lower crystallinity.