The FFD is a systematic approach used in research and experimentation to examine the effects of multiple factors, each at different levels, and understand their individual and combined effects on a particular response of interest. FFD provides a complete and detailed analysis of how each factor influences the response variable, enabling researchers to make informed decisions [
24]. In this study, the effect of three independent variables—temperature, extraction time and L/S ratio, as well as their interactions—on the total sugar contents was investigated using a full factorial design as a DOE. ANOVA was carried out to check the significance of the proposed models and to determine the most significant factors.
3.1. Variety Alig
Table 2 shows a coefficient of determination (R
2 = 0.992) close to 1, which means that 99.26% of the total variations in the data can be explained by the developed model (Equation (2)). The analysis of variance (ANOVA) revealed that the model was statistically significant (
p < 0.05), confirming its reliable prediction of TSC values. The model F-value of 53.75 further supports the model’s significance, suggesting that its performance was not related to random variation.
The marginal discrepancy of 1.85% between the R
2 and R
2 (adj) values suggests a reduced likelihood of non-significant terms impacting the model. Furthermore, the proximity of the adjusted R
2 coefficient (0.9741) and predicted R
2 (0.8818) confirms the strong predictive ability of the developed model.
Figure 1a also shows that the effects of first-order factors (temperature, time) and interaction factors (temperature, time) were significant (
p ≤ 0.05). For additional validation, a Pareto chart was generated, incorporating a reference line (t-value limit) to indicate the significance of each factor, thus ranking them in descending order. Factors crossing this reference line were considered theoretically significant. According to the obtained data, the main effects of time (X
2), temperature (X
1), and interaction factors, such as time–temperature (X
2 × X
1) and time–L/S ratio (X
2 × X
3), exerted a significant influence on the TSC in Alig extract, with the highest coefficients being −15.88, −10.63, −9.37, and −5.62, respectively. On the other hand, the other factors showed no significant effect on the response.
Figure 1b illustrates that the extraction time and temperature negatively influenced the TSC in date extract, while the L/S ratio (X
3) seemed to positively affect the response.
In addition to the main effects, the independent variables (X
1, X
2, X
3) may interact with each other to affect the response (TSC). For instance, extraction time and L/S ratio may interact to affect the TSC in Alig extract. At a short extraction time (20 min), the TSC increases from 519 to 540 mg/g of DP when the L/S ratio increases from 6 to 10 mL/g. However, at a prolonged extraction time (60 min), the TSC decreases from 521 to 512 mg/g of DP (
Figure S1 in Supplementary Materials). Therefore, the effect of the L/S ratio varies the response at two different levels of extraction time, suggesting an interaction between these two process parameters. Furthermore, the influence of extraction time on the response (TSC) differs at two different levels of temperature, confirming the significant interaction between these two factors. The TSC reaches its maximum quantity when the temperature and extraction time are both at low levels. Regarding the interaction between temperature and L/S ratio,
Figure S1 shows parallel lines, indicating that there is no interaction effect between these factors.
For more details, the 3D surface plots and 2D contour plots (
Figure 2) were used to determine the values of the response with different combinations of the independent variables (X
1, X
2, X
3) and to visualize the interactions between these variables. The contour lines or curves link points with constant response values, aiding in the selection of the best combinations that produce the desired response. Linear contour curves suggest the insignificance of the interaction term, while pronounced curvatures in the contours indicate a substantial and crucial interaction term [
25]. In our case, the contour curves display significant curvature (
Figure 2a), indicating that the interactions X
2 × X
1 and X
2 × X
3 are significant. However,
Figure 2b shows linear lines, which means no significant interaction for X
1 × X
3. Furthermore, the contour plot provides the best factor settings for a maximum TSC in the date extract. As shown in our results, the TSC increased from 475 mg/g of DP to 540 mg/g of DP as the extraction temperature and time varied from 40 to 55 °C and 20 to 40 min, respectively, which can be explained by the enhanced solubility of sugar particles [
26]. On the other hand, extending the extraction time beyond 40 min led to a rapid reduction in TSC, which can be explained by the degradation of sugar particles and formation of impurities like Maillard reaction products [
27].
The extraction process was also conducted by changing the L/S ratios from 6 to10 mL/g while maintaining the time and temperature at 20 min and 40 °C, respectively. As reported in
Figure 2c–f, the TSC was enhanced with a variation in the L/S ratio from 6 to 10 mL/g, leading to a decrease in solvent viscosity and dissolution of more sugar molecules in water. Therefore, the maximum TSC was set within the following conditions: 20 min, 10 mL/g, and 40 °C, which were considered for further validation.
To validate the predicted model, the ultrasound-assisted extraction of date sugar was carried out under the proposed conditions (40 °C, 20 min, and 10 mL/g). The obtained extract yielded a TSC of 608.710 mg glucose/g DP, which closely aligns with the predicted value indicated by the proposed model (537.750 mg glucose eq/g DP), as shown in
Figure 3. These results were obtained with a desirability of 96.53%.
3.2. Variety Kentichi
As shown in
Table 3, a high coefficient of determination (R
2 = 99.83%) with a low standard error (1.215) indicates that the model effectively explains the variation in the TSC values in the Kentichi extract. Consequently, the suggested model is expressed by Equation (3).
The analysis of variance (ANOVA) revealed that the model is highly significant (
p < 0.01), meaning that it effectively predicts the response values based on the extraction parameters (
Table 3). Additionally, the adjusted R
2 coefficient (0.9736) and predicted R
2 (0.9942) confirmed a robust correlation between the input variables and the studied response. The effects of first-order (L/S ratio, temperature) and two-way interactions (temperature time) were significant (
p ≤ 0.01). The Pareto chart (
Figure 4a) indicates that the L/S ratio (X
3), the interaction time temperature (X
2, X
1), and the extraction temperature were the most significant factors influencing the TSC in Kentichi extract.
Figure 4b reveals the high and positive effect of L/S ratio (X
3) on TSC in Kentichi extract, while temperature (X
1) and time (X
2) exhibited lower negative effects, indicating that increasing the temperature and extraction time leads to a decrease in the extracted sugar. The high effect of the L/S ratio on the response may be explained by the dry texture of the Kentichi variety, which needs more water to dissolve sugar particles and to enhance the matter transfer.
Furthermore, the independent variables (X
1, X
2, X
3) may interact with each other to affect the response (TSC). At a short extraction time (20 min), the TSC decreased from 511.2 to 500 mg/g of DP when the temperature increased from 40 to 75 °C. However, after 60 min of extraction time, the TSC increased from 490 to 500.1 mg/g of DP (
Figure S2 in Supplementary Materials). Therefore, the effect of extraction time on the TSC depends on the level of the extraction temperature (min or max), indicating the presence of interaction between these two factors. For the interaction between time and L/S ratio, the results reveal that L/S ratio inversely affects the output at two different extraction times, which confirms the interaction between these two parameters. Parallel lines disprove any interaction between temperature and L/S ratio.
The 3D surface plots and 2D contour plots (
Figure 5) were also utilized to highlight the interactions between the studied variables.
Figure 5a,b demonstrate a strong curve, indicating a notable interaction between temperature and time, while
Figure 5e,f reveal softer curves, meaning a weaker interaction between time and L/S ratio. The contour plot depicting temperature versus L/S ratio displays linear lines, demonstrating there was no significant interaction between temperature and L/S ratio. The TSC dropped drastically as the extraction temperature and time exceeded 40 °C and 20 min, respectively (
Figure 5a,b). The optimal extraction temperature and time resulting in the highest TSC were around 40 °C and 20 min, respectively. In addition, the sugar extraction was also performed by changing the L/S ratios (6 to10 mL/g) while keeping the temperature and extraction time constant. The increase in the L/S ratio from 6 to 10 mL/g ameliorated the TSC in Kentichi extract (
Figure 5c–f). Therefore, the maximum TSC was registered under the following conditions: 40 °C, 20 min, and 10 mL/g, which were considered for further validation.
To validate the proposed model, the extraction procedure was repeated in triplicate under the optimum conditions described above (40 °C, 20 min, 10 mL/g). The experimental value registered was 507.64 mg glucose/g DP, which is consistent with the predicted value given by the model (539.577 mg glucose eq/g DP), as shown in
Figure 6. This result was obtained with a desirability of 100%.
In conclusion, the highest extraction efficiency was achieved for both date varieties under the same following conditions: 40 °C extraction temperature, 20 min extraction time, and 10 mL/g liquid-to-solid ratio. These findings are consistent with previous studies. Alyammahi et al. [
16] suggested an extraction temperature of 60 °C, an extraction time of 30 min, and an L/S ratio of 7.6 mL/g to obtain the optimum sugar content from the Sukkari date variety. According to Entezari et al. [
28], a low temperature (15 °C) and a high water/fruit ratio (9/1) and ultrasonic intensity provide the best sugar content from date fruit. Lin et al. [
29], who studied the extraction of polysaccharide from
Ziziphus jujube Mill var
spinosa seeds (ZSS) using ultrasound-assisted extraction (UAE), confirmed the significant effects of extraction time and sonication power on extraction yield. They reported the following optimizing conditions: 52.5 °C, 21.2 min, 134.9 W, and L/S ratio of 26.3 mL/g. In our study, a maximum sugar content was obtained using less energy (only 40 °C) and a short extraction time (20 min), leading to a reduction in the processing cost compared to conventional extraction methods.
3.3. Characterization of Optimized Date Extracts
After optimizing the sugar extraction process, date extracts of the studied varieties were assessed for their sugar and mineral profiles and their composition in terms of protein, fats, and polyphenols. The results are summarized in
Table 4.
Sugars were the most abundant component in both varieties, ranging from 50.79% to 60.2% in Kentichi and Alig, respectively. Alig extract showed a higher sugar percentage than Kentichi extract, with a
p < 0.001. Our results are lower compared to those obtained by AlYammahi et al. [
16], with 812 mg/g of DP for the Sukkari variety, and those provided by Messadi et al. [
9], with 679.1 mg/g of DP for the Kentichi variety. The variations in sugar content are attributed to the fruit variety and grade. In fact, in the present study, we used a very low quality of date fruit which is usually considered as waste. The protein content was quite low for both varieties, with values of 1.39 g/l and 1.07 g/l for Alig and Kentichi, respectively, with a significant difference between varieties (
p < 0.001). A negligible amount of fat was registered, with values of 0.001% and 0.002% for the varieties Alig and Kentichi, respectively. Furthermore, a good level of polyphenols was found, with amounts of 333 mg eq GAE/100g and 254 mg eq GAE/100g for the Alig and Kentichi varieties, respectively. These amounts are lower than those provided by Saafi et al. [
30] and higher than those reported by Lajnef et al. [
31], who found a total phenolic content of 138.97 mg eq GAE/100 g in date syrup. The differences observed in the nutritional composition of both varieties of dates when compared to the literature could be explained by several factors, such as the variety, date cultivar, geographic origin, harvest season, solvent used, and extraction method (process parameters) [
32].
For a more detailed view, the sugar profiles of date extracts were screened via high-performance liquid chromatography with refraction index detection, and the results are displayed in
Table 5. They showed that the amount of reducing sugars was significantly higher in Kentichi extract than in Alig extract (
p < 0.001). Moreover, the composition of simple sugars revealed some differences between the studied varieties of dates. For example, in the Alig extract, fructose (35.4%) and glucose (30.3%) were the predominant sugars, with the sucrose content being minimal (0.2%). However, the sucrose content was significantly higher in the Kentichi extract (22.8%) compared to Alig. Additionally, other simple sugars were present, including maltose, maltotriose, and lactose, with amounts of 3.8%. 2.1%, and 1%, respectively. Good levels of fructose (24%) and glucose (17.9%) were also registered in the Kentichi extract. These differences in the sugar composition may be related to the fruit texture. Typically, soft date varieties are rich in reducing sugars (glucose and fructose), with a low amount of sucrose, whereas dry date varieties contain much higher contents of sucrose [
33]. These results are consistent with the value given by AlYammahi et al. [
16] in their study of a common date variety (Degla-beida). They reported a high sucrose concentration in the Sukkari date variety (~85 mg/mL). In another study, Messadi et al. [
9] indicated the predominance of sucrose (72%) over glucose (14.26%) and fructose (13.74%) in Kentichi extract. Notably, maltose and lactose were no longer present in Kentichi extract after 2 h of extraction time. In our case, the presence of maltose, maltotriose, and lactose may be explained by the short extraction time (20 min) employed. These findings demonstrate that high temperature and longer extraction time alter the sugar composition of dates.
Date fruits are further recognized for their high content of minerals. Concentrations of essential minerals such as K, Mg, Ca, Na, Fe, and Zn for the studied date extracts were examined by using an ICP-MS iCAP-Q spectrometer and are provided in
Table 6. The results show that potassium (K) was the major mineral element present in both date extracts, with values of 363.79 mg/100 g of DP and 349.82 mg/100 g of DP, respectively, for the varieties Alig and Kentichi. The macroelement profile reveals high Mg, Ca, and Na concentrations in the ranges of 40.82–43.01 mg/100 g of DP, 38.02–38.32 mg/100 g of DP, and 9.60–9.86 mg/100 g of DP, respectively, for both varieties of date fruit. Iron and zinc essential microelements were found in high amounts ranging between 1.56–1.67 mg/100 g of DP and 0.95–1.09 mg/100 g of DP, respectively, for the studied varieties. These results are consistent with the values presented by Tang et al. [
32], who studied the chemical compositions of various date fruit varieties.
Date extract provides several benefits due to its rich nutritional profile and antioxidant content, and it could be a better option than commercially refined sugar. Several studies have shown the ability to use date sugar extract in many sectors, particularly in the food industry. A study carried out by Lajnef et al. [
31] reported the use of date syrup as a sugar replacement in sponge cake. Previous results have shown that the final product has a texture profile which is very close to the control, with high antioxidant activity. Moreover, Rangaraj et al. [
34] demonstrated that the incorporation of date fruit syrup waste extract in gelatin films delays lipid peroxidation and extends the shelf life of canola oil. AlYammahi et al. [
35] developed a fortified camel milk powder enriched with date sugar extract using the spray-drying technique. The final product has a high nutritional value, with a high content of poly- and mono-unsaturated fats, a low cholesterol content, and good thermal stability, making it a value-added food. Furthermore, date extract could be used as a potential substrate for microbial fermentations. Salah-Tazdaït et al. [
36] examined the use of crude juice from dates as an additional carbon source for malathion degradation by
Stenotrophomonas maltophilia. Previous results have indicated a significant increase in biomass growth, leading to the improvement of malathion biodegradability. Also, Elsanhoty et al. [
37] investigated the production of carotenoids by
Lactobacillus plantarum QS3 using date syrup as a source of sugar. The obtained results indicated that 5% of date syrup produces 16.21 mg/kg dry cell of carotenoids. A similar study conducted by Acourene and Ammouche. [
38] demonstrated that date waste is a potential substrate for the synthesis of several fermentation products, including ethanol, citric acid, and α-amylase. Thus, we have observed that agro-industrial by-products such as date waste may serve as great ingredients for various biotechnological applications. Furthermore, date extract was successfully used as a carbon source to enhance biopolymer production yield. For example, Lotfiman et al. [
39] revealed a 68% increase in bacterial cellulose production yield using 3% (
w/
v) dates in the medium after 8 days of incubation compared to the standard medium.
Therefore, striving to expand the circular economy concept helps to ensure a safe environment and a high quality of life for humankind.
3.4. FT-IR Analysis
The FTIR spectra of date fruit, solid residue, and liquid extract for the studied date varieties are presented in
Figure 7. The date fruit spectrum (AF, KF) showed a large band at 3255 cm
−1 attributed to hydroxyl groups (–OH). The characteristic band at 2942 cm
−1 corresponded to the CHO group. The typical absorption band at 1640 cm
−1 was associated with the vibration of C=C aromatic and COO
–. Several bands were detected at the fingerprint region between 1500 and 450 cm
−1, corresponding to the carboxylic group (C=O) of carbohydrates, acids, ketones, and aldehydes. A strong absorbance was registered around 1170 to 1050 cm
−1, which was attributed to polysaccharides. However, in the date residue spectrum, no significant peaks were detected for either variety of date, indicating that sugar molecules and other phytochemicals were successfully extracted from the date fruit. Additionally, the spectral pattern did not change after sonication, indicating that the chemical composition of the studied material was not altered. This suggests that sonication treatment is a potential tool for molecule recuperation with less damage. Therefore, ultrasound extraction demonstrated a strong performance in terms of extracting sugar molecules from date fruits while preserving the chemical integrity. These results are in line with those provided by AlYammahi et al. [
16], who performed sugar extraction using another variety of dates.
3.5. Scanning Electron Microscopy (SEM)
The surface morphology of Alig fruit (AF), Kentichi fruit (KF), Alig residue (AR), and Kentichi residue (KR) was assessed using SEM and is displayed in
Figure 8 and
Figure 9. The untreated date samples (AF, KF) showed firm and compact surfaces (
Figure 8a,c,e and
Figure 9a,c,e). However, much damage was recorded with the treated ones (AR, KR) using different magnifications. As shown in
Figure 8b,d,f and
Figure 9b,d,f, many holes and cracks with sharp edges were generated on the surfaces of both varieties of date. These surface changes are explained by the strong effect of the acoustic cavitation generated by the ultrasonic wave, which attacked the cell walls and released the internal content [
40]. In fact, the ultrasound waves navigated through the liquid solvent while producing acoustic cavitation, which caused particle fragmentation, cell disruption, and sonoporation. Therefore, the permeability of cell walls was improved, and the extraction yield was boosted [
41]. Our results were in agreement with those provided by Santos et al. [
42] and Dias et al. [
43], who showed that the use of ultrasound waves for vegetal materials improved the extraction yield due to the lateral cracks formed, which led to the release of internal particles.
In conclusion, the ultrasound extraction technique represents a promising alternative to conventional methods for the isolation of valuable nutrients from vegetal matrices [
44].