3.1. Moisture Content and Water Activity
While “NP” started to dry with a moisture content of 93.8% wet basis (w.b.), all bitter gourds dipped at different °Brix values for 10 h started drying between 52.1% and 63.9% (w.b.) moisture content. Although “S70” started the drying process at the lowest moisture content, “G50” was the first to go below 10% moisture level after 180 min. In addition, “G50” started the drying process with the second highest moisture content. “G70” started at the highest level and reached an equilibrium moisture content with the highest moisture content as a result of drying. “S50” and “G60” started to dry at close moisture contents: “G60” lost moisture more quickly at the beginning, but after 135 min, “S50” started to lose moisture faster. While only “G50”, “NP”, “S70”, and “S50” dropped below 10% (w.b.) moisture content, they exceeded this limit at drying times of 180, 250, 255, and 345 min, respectively (
Figure 3).
All of the bitter gourd dipped in different Brix values for 20 h remained between 41.4% and 62.4% (w.b.) moisture content at the beginning of drying. While all pre-treatment trials with “S” were below 10% (w.b.), the closest “G60” pre-treatment trial approached here in “G” application and reached this level in approximately 360 min. All samples treated with “S” reached 10% (w.b.) levels in the range of approximately 135–145 min. “S70” started the drying process from the lowest moisture content in the 20 h process, as in the 10 h dipping time pre-treatment. Then, while “S50” and “S60” caught the lowest level start, when it went below 10% level, this situation was reversed, and these two trials lost moisture faster. In “G” applications, “G50” started the drying process with the highest moisture content, followed by “G60” and “G70”, respectively. At the end of drying, while “G70” remained at the highest moisture content, “G50” did not decrease to 10% (w.b.) moisture content. Among all dipping pre-treatments, the trial in which the moisture content levels of 10% (w.b.) was decreased in the shortest time was seen as the application in which 20 h dipping time was made (
Figure 4).
All bitter gourds, which were kept at different °Brix values for 30 h, started to dry between 49.4% and 64.8% (w.b.) moisture content values. “S70” started drying with the lowest moisture content as in all other pre-treatment times. Then, “S60” was seen at the lowest level. Although “S50” and “S60” started to dry at a higher moisture content level than “S70”, “S60” at the 70th minute and “S50” at the 125th minute passed “S70” and went below 10% (w.b.) moisture content at approximately 180 min. After these, the “S70” exceeded the same level in the 300th minute. While “G50” started at the highest moisture content level, the drying curve changed at the 105th minute, and “G70” completed drying at the highest moisture content level as in all other dipping times. “G50” and “G60” showed the same curve after the 255th minute. Although “S50” started to dry at the highest moisture content level after “G50”, it caught a fast drying curve and decreased below 10% (w.b.) moisture content in the shortest time (
Figure 5).
Due to its low molecular weight, sucrose is one of the osmotic agents. Sucrose was used as an osmotic agent in previous osmotic pre-treatment studies. Sucrose can readily diffuse and permeate into cells and tissues; thus, it may protect the quality of vegetables and fruits by inhibiting many chemical processes including enzymatic browning and oxidation [
16,
35]. In one of previous research work, Ispir and Togrul [
8] investigated the effects of four different osmotic agents on the mass transfer of osmotically pre-treated apricots and discovered that sugar caused the water loss. Therefore, the researchers thought that sugar was one of the best osmotic agents. Regarding moisture contents of bitter gourd slices that are affected at different °Brix values of sugar and grape molasses, it is revealed that sugar is more effective than grape molasses for the bitter gourd slices that reach less than 10% (w.b.) moisture content in the shortest time. On the other hand, it is assumed that the concentration of the osmotic solution and osmotic pre-treatment duration affects the moisture content of an osmotically pre-treated vegetable and fruit sample. It was discovered that removing moisture from pineapples was successful when using sugar syrup at 70 °Brix [
16]. Pre-treatments like blanching also improve mass and heat transmission as well as enhancing qualities of fruits and vegetables.
When the samples were evaluated in terms of a
w, a statistically significant difference was observed between the samples with “S” and “G” pre-treatment, and it was determined that all samples treated with “S” were higher. However, all of the “S” treated samples had lower a
w levels than the fresh sample. After the fresh trial, the highest a
w values were observed in the 50 and 70 °Brix solutions samples at 10 and 20 h dipping times. Among all the pre-treated samples, the lowest a
w value was seen in the 30h50Bx sample, while it was in this group in all samples in the 30 h group. The a
w values were higher than “NP” in all “S” trials. While all “G” samples do not show any statistical difference in themselves, this situation is different in “S” samples. In the “G” trials, “NP” and all other pre-treated samples were at a lower level than the “Fresh” sample and were statistically in the same group (
Figure 6). For the stability and quality assurance of dried foods, it was reported that water activity was more crucial than moisture content [
13]. As the amount of water contained in foods an increase, the water activity value also increases. The occurrence and speed of many chemical reactions, such as microbial growth, are related to water activity. To extend the shelf life of vegetables and fruits, it may be necessary to ensure they are dry. Fruits and vegetables that are osmotically air-dried become unstable when stored. In the effort to inactivate enzymes, the blanching procedure was used after osmotic treatment but before further processing [
16]. Nyangena et al. [
36] showed that pre-treatment before drying was a necessity to contribute to lower the water activity content of mango chips. It is thought that lower water activity slows down the activity of bacteria, enzymes, yeasts, and molds [
37].
3.2. Mathematical Modeling of Drying Processes
Thirteen different types of mathematical models for bitter gourd were made. The gourd was sliced into 8-mm-thick slices at 70 °C drying air temperature and dipped in different °Brix values for different times and then dried. The most suitable models were examined. These analyses were evaluated separately for sugar (S) (
Table 4) and grape molasses (G) (
Table 5).
As a result of all “S” pre-treatments, the “Diffusion Approach”, “Logarithmic”, and “Midilli et al.” models gave the highest R
2 and lowest X
2—RMSE values, respectively. In the “NP” pre-treatment, the highest R
2 (0.999) was seen in the “Midilli et al.” model, followed by the “Two-term”, “Modified Henderson and Pabis” and “Verma et al.” models with a value of 0.9976. The samples with “S” pre-treatment are the same as the model with the lowest X
2 value of “NP”. The lowest RMSE values of “NP” was seen in the “Midilli et al.”, “Verma et al.”, and “Two-term” models, respectively. Among the samples with “S” pre-treatment, the highest R
2 was seen with a value of 0.9999 in the 10h60Bx, 20h60Bx, and 30h70Bx samples in the “Diffusion approach” model. The smallest X
2 is again in the same model and was calculated as 1.35 × 10
−6 in the 30h70Bx trial. The minimum RMSE was calculated as 0.0012 in the same model, in the 30760Bx trial (
Table 4).
Following all preprocessing treatments of “G”, the “Diffusion Approach”, “Logarithmic”, and “Midilli et al.” models yielded the highest R
2 values while demonstrating the lowest X
2—RMSE values, respectively. The same results were seen in the “NP” process. The samples with “G” pre-treatment and the model with the lowest X
2 value of “NP” are the same. Among the “G” pre-treated samples, the highest R
2 “Diffusion approach” model was observed with a value of 0.9999 in the 10h50Bx, 10h60Bx, and 30h70Bx samples. In addition, all R
2 values in the same model are between 0.9985 and 0.9999 in other trials. While the smallest X
2 was seen in the same model in all pre-treatments, the same value was seen in the “Logarithmic” model, with the value of 4.60 × 10
−5 in the 10h60Bx trial. The smallest RMSE was calculated as 0.0012 in the “Diffusion approach” model, in the 30h70Bx trial (
Table 5).
Shahari et al. [
38] carried out a study on the mathematical modeling of cucumber drying, which is also from the same family as bitter gourd, at a drying air temperature of 50 °C. In the study, the ”Newton”, ”Page”, ”Modified Page”, ”Henderson and Pabis”, ”Logarithmic”, ”Wang and Singh”, and “Midilli et al.” models were tested, and the most suitable model was chosen, since the “Logarithmic” model gave the highest R
2, lowest RMSE, and SSE (sum of square error) values. da Cunha et al. [
39] dried the melon at 60 °C drying air temperature by applying four different pre-treatments as dipping, dipping with ultrasound, with vacuum and with ultrasound and vacuum. As a result of drying, the “Single exponential”, “Henderson and Pabis”, “Logarithmic”, “Two-term”, and “Wang and Singh” models were tried, and “Two-term exponential” gave the best model result.
During drying the immature bitter gourd with a slice thickness of 5–7 mm with a solar dryer, 10 different models, ”Lewis”, ”Page”, ”Henderson and Pabis”, ”Two-term”, ”Logarithmic”, ”Wang and Singh”, ”Two-term exponential”, ”Verma et al.”, ”Approximation of diffusion”, and ”Midilli–Kucuk” were tried to make the mathematical modeling. As a result of the study, the “Two-term” and “Midilli–Kucuk” models gave the highest correlation coefficient R
2 and the lowest RMSE values [
40].
After the 13 models were tested, the model constants of the three models that gave the best results for the “S” pre-treated samples were examined. In all three models, the highest “a” value was observed in the 30h50Bx trial, while the “k” value remained at the lowest level in the same trial. In a similar contrast, in the 20h70Bx trial where the ”a” value was the minimum, it was observed that the ”k” value was the maximum. Among these three models, the maximum ”a” value was 0.6430 in 30h50Bx; the minimum ”a” value was 0.3480 in 20h70Bx in trials in the ”Diffusion approach” model. The maximum ”k” value was 0.0268 in the 20h70Bx trial, in the ”Midilli et al.” model, and the minimum ”k” value was 0.0121 in the 30h50Bx trial in the ”Diffusion approach” model (
Table 6).
Among the samples dried by dipping “G”, the best models were found to be “Diffusion approach”, “Logarithmic”, and “Midilli et al.”, as was the case with “S” immersed samples. In the “Diffusion approach”/“Logarithmic” models, the maximum/minimum “a” values were found to be 0.5919/0.4154 and 0.6092/0.4398 in the 10h50Bx/30h70Bx trials, respectively. In the same models, the maximum/minimum “k” values were found to be 0.0194/0.0185 and 0.0145/0.0148, respectively, in the 20h60Bx/30h50Bx trials. While the maximum and minimum values were seen in similar trials in all three models in the samples with “S” pre-treatment, it was observed that the “Midilli et al.” model did not give similar results compared to the other two models in the samples immersed with “G”. In the “Midilli et al.” model, the minimum “a” and “k” values were found in the trials of 10h60Bx (1.0015) and 30h50Bx (0.0179), respectively. The maximum “a” and “k” values were found in the 30h50Bx (1.0042) and 10h50Bx (0.0316) trials, respectively (
Table 7).
3.3. Color Measurement
Color is a crucial aspect of fruits and vegetables since people choose which ones to eat and buy based on their color. Color alterations can occur as a result of chemical reactions such as enzymatic browning and non-enzymatic browning, which result in the production of brown pigmentation and are facilitated by enzymes like polyphenol oxidase or peroxidase [
35]. Among the “S” trials, it was observed that the highest ΔE value was in the 30h60Bx, followed by the 20h50Bx trial. All “S” trials ΔE value are higher than “NP”. ΔE value is higher in all “G” trials than “S” trials. In the “G” trials, a significant decrease was observed in the ΔE value due to the increase in the Brix value in the samples dried in 10 h of dipping time. No such difference was observed in the 20 h application. In the 30 h dipping process, it was the opposite compared to the 10 h dipping process, and the increase in the Brix value increased the ΔE value. The highest ΔE value was seen in the 30h70Bx trial, while the lowest was in the 20h60Bx trial (
Figure 7). Because of the increase in pigment density throughout osmotic pre-treatment and the drying process, Falade et al. [
41] noticed that watermelon that had been osmotically dried had a deeper hue. The variation of concentration of osmotic pre-treatment solution may cause differences in the appearance of vegetables and fruits. In a previous study, Kowalski and Mierzwa [
42] demonstrated that the color change took place between 20% and 60% sucrose solutions of osmotically pre-treated carrot samples. Similarly, the results of the present study revealed that alterations in the osmotic pre-treatment solution concentration make differences in the total color changes. Because grape syrup solutions are rich in monosaccharides (glucose and fructose), they have a greater impact on color indices when used as osmotic solutions. Changes in the color index during the process are probably influenced by the colored pigments of osmotic solutions (grape and mulberry syrups), pigmented chemical interactions such as Milliard’s reaction between sugar and protein, and production of Melanoidin [
43]. The ΔE value represents how much the dried product’s color changed overall. The finest samples in terms of visual appeal are those with low color change values (ΔE). The untreated samples showed minimum color change. In comparison to grape molasses pre-treated samples, those that had been pre-treated with sucrose exhibited the fewest color alterations, after the untreated samples. The interactions and synthesis of color pigments are actively influenced by the chemical composition of glucose and fructose. Due to the high amount of monosaccharides including glucose and fructose that they contain, osmotic solutions like grape and mulberry syrups have an impact on color changes. Therefore, it is thought that it may be the reason why the color change in the “G” trials is greater than in the “S” trials [
43].
3.4. Total Phenolic Content and Antioxidant Activity
One of the most significant classes of compounds in plants is the phenolic compounds [
44]. Regarding total phenolic contents, differences were examined by applying statistical analyses after bitter gourd drying, in which different applications were conducted. Differences were examined by applying statistical analyses after bitter gourd drying, in which different applications were made. From the phenolic content of view, all samples treated with “S” showed a decrease, while all samples treated with “G” were higher than those with “S”, and this difference is statistically significant. One of the fruits with the highest concentrations of phenolic compounds is grapes; as a result of their high phenolic content, these fruits have high antioxidant activity [
45]. Therefore, grape molasses is rich in phenolics and antioxidants, and it is assumed that grape molasses dipping samples demonstrated higher total phenolics and antioxidants than sugar dipping samples. Despite the different Brix and dipping times, there was no statistically significant difference between all “S” treated samples. While the highest phenolic content 50 °Brix value was observed in the samples that were immersed for 10 h, the “G” mixture and the 60 and 70 °Brix mixtures decreased gradually; this decrease is statistically significant. It was observed that dipping pre-treatment at different Brix values in the 20 h treatment, the “G” mixture did not cause any change in the samples in terms of phenolic content. In the 30 h and “G” mixture, 60 °Brix shows the highest phenolic content value, while there is no difference between the other Brix values. In addition, bitter gourd dried by dipping in a mixture of “G” for 30 h and 60 °Brix, although not statistically significant, showed the highest phenolic content value, outstripping the “Fresh” and “NP” samples. While “NP” and 10h50Bx samples followed the 30h60Bx, there was no statistical difference between these three highest samples (
Figure 8). The greater antioxidant activities were determined at non-pre-treated bitter gourd samples. The differences in antioxidant activity of bitter gourd slices at different dipping solutions showed different responses to dipping time treatment. Increased antioxidant activity was determined in high-Brix-treated grape molasses treatment samples and in the 20 h dipping time, which means that increasing dipping time ends with an increase in phenolic content. In a similar trend to the present study, Dermesonlouoglou et al. [
46] demonstrated the highest total phenolic content with increased osmotic pre-treatment time in the goji berry. The increase in antioxidant activities of bitter gourd samples might be due to the increase in phenolic content and variation of polyphenols in the matrix after dipping time and thermal processing as reported by Kim et al. [
47]. It is assumed that polyphenols and antioxidant activity are highly linked to each other. Therefore, it is possible to say that the increasing total phenolic content of bitter gourd slices is a consequence of thermally treated sugar and grape molasses dipping procedures.
The antioxidant activity values of all samples subjected to both “S” and “G” pre-treatment at different dipping times were lower than the “NP” group. A decrease in the antioxidant activity level was observed due to the increase in the Brix value in the “G” pre-treatment at 10 h dipping time. Although an increase was observed due to the increase in the Brix value in the 20 h dipping time in the “G” pre-treatment, there was no statistical difference between the 60 and 70 °Brix values. The increase in antioxidant activity was probably caused by the increased Brix of grape molasses solution when the dipping time was increased from 10 to 20 h. The lowest antioxidant activity in the “S” trial was seen in the 20h50Bx trial followed by the 10h60Bx trial. In the “G” pre-treatment trial, the lowest antioxidant activity was observed at 10h70Bx, and this trial was in the lowest group statistically. No statistical difference was observed in different Brix values in both “S” and “G” trials in the 30 h dipping pre-treatment. Only in the 10h60Bx and 20h50Bx trials was there a statistical difference in terms of antioxidant activity between the “S” and “G” pre-treatments (
Figure 9). The significant release of soluble antioxidant substances through the osmotic pre-treatment process is assumed to be the cause of the reduced antioxidant value in osmotic pre-treatment bitter gourd slices compared to non-pre-treated bitter gourd slices. It is believed that grape molasses can boost antioxidant activity while preventing the degradation of phytochemicals [
45].
3.5. Total Carotenoid Content
Carotenoids are delicate substances which are easily degraded by acid, light, and high temperatures [
48]. When the dried samples were examined in terms of carotenoid content, it was determined that the “S” treated samples showed higher results than the “G” treated samples in all trial repetitions, apart from 10 h 60 °Brix and 30 h 50 °Brix, and it was statistically significant. However, the carotenoid content value is higher in “Fresh” and “NP” samples than in pre-treated samples. It is thought that the cell structure is affected by the process and carotenoids are leached during the osmotic pre-treatment. Since carotenoids are rich in conjugated double-bond structures, they are easily degraded or isomerized during heated air drying [
49]. The dipping time and different Brix values between the “S” treated samples did not cause any difference. On the other hand, 10 h 60 °Brix and 30 h 50 °Brix trials were statistically higher in “G” samples than all other “G” samples. Although there is no statistically significant difference, the highest carotenoid content values in dried bitter gourd samples applied “S” and “G” were observed in 10 h 50 °Brix and 10 h 60 °Brix trials, respectively (
Figure 10). Luchese et al. [
50] reported that osmotic solution concentration did not significantly affect the carotenoid content of osmotic pre-treatment of physalis fruits. The same researchers indicated that the higher the temperature during the osmotic pre-treatment process caused the higher loss of carotenoids from products. It was also reported that regardless of the osmotic solution concentration, physalis fruits which was subjected to a 10 h osmotic pre-treatment process had no carotenoid loss. In the current study, non-pre-treated and fresh samples of bitter gourd slices showed higher values of the carotenoid content than the bitter gourd slices pre-treated with various concentrations of sugar and grape molasses osmotic solutions. Sanjinez-Argandona et al. [
51] found a reduction on the total carotenoid content of guava which was subjected to air drying after osmotic pre-treatment. It was also reported that the loss of carotenoid was much greater when the guava samples were subjected to a hot-air drying process without osmotic pre-treatment. It is hypothesized that the loss of carotenoids is triggered by high drying temperatures, increased exposure to oxygen, and the low relative humidity of the drying air as they all accelerate the decomposition rate [
50].