Developing ultrasound-assisted hot-air and infrared drying technology for sweet potatoes

Highlights • US assisted drying with two drying techniques employed for the first time for SP.• Hii model found fit for predicting the drying kinetics in HAD and IR.• Lower US frequencies preserved the nutritional features in dried SP.• The developed drying protocol helpful in energy consumption.


Introduction
Sweet potato (Ipomoea batatas L.) is technically a perennial vegetable vine crop, but it is typically grown as an annual due to its warm-weather requirements. Since it is a seasonal crop that must be utilized within a specific time frame after harvesting, it has a short storage period with high moisture content (MC) that encourages microbial activity, leading to deterioration. The convective hot-air drying process is frequently used for preserving sweet potatoes. It is rich in amino acids, β-carotene, starch, cellulose, vitamin C, and other nutrients. Sweet potatoes have medicinal benefits such as maintaining the body's acid-base balance, preventing cancer, and lowering blood fats. Raw or processed sweet potatoes can be used in bakery products and snacks and consumed as a staple food [1]. New non-destructive technologies are always a concern for scientists as the world's population grows and non-renewable energy sources become scarcer. The factors commonly considered for implementing new technologies include minimizing energy consumption, enhancing product quality, increasing production rates, and reducing processing time [2]. Moisture reduction or drying (through simultaneous mass and heat transfer) is usually employed to extend the shelflife of vegetables and fruits, reduce postharvest losses, and maintain quality [3]. Numerous techniques are used to dry food materials, including hot-air, solar, oven, microwaves, or infrared drying. One of the oldest and most frequently employed methods is hot-air drying which involves simultaneous heat and mass transfer and phase change, making it an energy-intensive approach [4][5][6]. The necessity of high-quality fastdried foods leads to a renewed interest in drying operations. In addition, there is an increased demand to combine non-destructive methods with different drying technologies to boost efficiency, and minimal quality reduction has gained attention. Application of pretreatments (including ultrasonic wave, blanching, microwave) is the recent addition involving mass transfer during air drying. Wherein ultrasonic waves technique, a non-thermal approach, is recently attracting the researcher's attention as an effective pretreatment for food drying and preserving the physical and sensory features of the end-product [3,5,7]. This approach is used in a variety of ways. For example, a long-range sound wave is transmitted where; M = moisture content at every time intervals, M e = equilibrium water content and M 0 = initial water content.
Eq. (2) was used to determine the DR of all samples at a given point.
where t 1 and t 2 denote the drying times (min), and Mt 1 and Mt 2 represent the samples' moisture contents (kg water/kg dry matter).

Analysis of drying data through mathematical modeling
To characterize the drying behavior of ultrasound pretreated samples, the drying kinetics of the samples were measured using various thin layer models (Supplementary Table-S1).

Analysis of non-linear regression
The non-linear regression analysis was performed using Origin 9.2. The excellent fit between modeling and the experimental data was estimated using four statistical functions (R 2 , RMSE, RSS, and reduced χ 2 ). Higher R 2 and lower RSS, RMSE, and χ 2 values indicate a better model fit [16].

Activation energy (E a ) & thermodynamics analysis
The activation energy (E a , KJ/mol) was estimated through Arrhenius Eq. (3).

Total energy consumption
An electric energy meter calculated every sample's total energy consumption (TEP) at various RH conditions calculated by the meter for electric energy (correctness of 0.01 kWh). The specific energy consumption (SEP) (the energy needed to dry 1 kg of sweet potato slices) was measured by Eq. (7) [17].
whereE s = ESP, E t =TEP, andW w = initial weight of samples.

Quality parameters 2.7.1. Analysis of total phenols (TPC) and flavonoid (TFC)
The TPC (mg.GAE.g − 1 ) was measured by our previously published method of dried sweet potatoes [18]. 2.5 mL of 0.2 N Folin-phenol Ciocelteu's reagent and 2 mL of 7.5 % Na 2 CO 3 were added to a 0.5 mL extract. The mixture was incubated at room temperature in the dark for 30 min. The sample absorbance was evaluated at 765 nm using a UVspectrophotometer (Tu-1810; Universal Instrument Purkinje Co., ltd., Beijing, China). For TFC, 1 mL of extract was put in a volumetric glass with a capacity of 10 mL. The volumetric glass was then filled with clean water until it held 5 and 0.3 mL NaNO 2 (5 %). After 5 min, 0.3 mL AlCl 3 (10 %) was added. After 6 min, 2 mL NaOH (1 M) was added, and the total volume of the solution was reduced to 10 mL by adding clean water. The resulting solution was thoroughly mixed, and the absorbance at 510 nm was measured. The TFC was examined by aluminum chloride assay for Rutin equivalent (mg.RU.g-1) [19].

Vitamin-C analysis by HPLC
Vitamin-C content (mg.100.g − 1 ) was evaluated using HPLC method [20]. 80 mL of 3 % meta-phosphoric acid solution was added to 5 g of sample, placed in a shaker for 10 min, and then the volume was made up to 100 mL using 3 % meta-phosphoric acid solution. Finally, the filtrate was shifted to HPLC vials. The detector's wavelength was 254 nm.

Total carotenoids
Total carotenoids were inspected in all samples by Tang et al. [21]. In brief, a 0.5 g sweet potato sample was extracted in triplicate using 5 mL of ethanolic butylated hydroxyl toluene (ethanol/BHT − 100:1, v/w) for carotenoid separation and release. The mixture was then mixed well before being put in a water bath at 85 • C for 5 min. After that, 0.5 mL of 80 % KOH was added for saponification and thoroughly vortexed before returning to the 85 • C water bath for 10 min. After cooling in an ice bath, the mixture was added to 3 mL of cold deionized water. N-Hexane (3 mL) was added to the mixture before centrifugation at 7500 rpm for 5 min to separate the two layers. The yellow top layer was transferred and collected. This treatment was repeated four times until the top layers became colorless. The samples were measured at 450 nm and 503 nm wavelengths against a blank of hexane. Total carotenoids (µg. g − 1 ) were computed by following Eq. (8).

β-Carotene determination
It was investigated by following the Jatoi et al. [22] method with no changes. Here, a 10 mL acetone-hexane (4:6; v/v) combination was added to a methanolic extract (100 mg sample), vortexed for 1 min, and filtered using Whatman No. 4 filter paper. The ultimate volume was determined to be 10 mL. Optical absorbance was observed at 453 nm, 505 nm, and 663 nm. The results are presented as µg.β-carotene g. -1 Fw (g.g − 1 Fw). The following equation was used to calculate the β-carotene levels:

Extraction of enzymes
As described by Sarpong et al. [16], the enzyme extraction was performed with no modifications. 5 g of dried sweet potato slices were powdered and combined with a 25 mL extraction solution (0.2 M phosphate buffer at pH 6.5, 1 % (v/v) Triton X-100, and 4 % polyvinyl polypyrrolidone). This was continually mixed for 3 min and maintained at 4oC for 4 h. The mixture was centrifuged for 10 min at 4 • C at 10,000 rpm, and the supernatant was collected and filtered through a 0.45 m filter membrane before being tested as crude enzyme extract.

Polyphenol Oxidase (PPO) activity
The PPO was measured in the same way as performed by Sarpong et al. [16]. The crude enzyme extract was evaluated by adding 0.5 mL to 200 l (0.1 M) catechol and 1.5 mL to 0.1 M sodium phosphate buffer (pH 7.0). The absorbance was measured continuously at 420 nm for 5 min. The sample prepared using an enzyme-free mixture solution was termed as blank.

Peroxidase (POD) activity
It was determined at 470 nm spectrophotometry by Zhang et al. [23] method without any modification. 0.15 mL of 4 % (v/v) guaiacol, 0.15 mL of 1 % (v/v) H 2 O 2 , 2.66 mL of 0.1 M phosphate buffer pH 7, and 40 µl of crude enzyme extract were used in the process. The solution mixture was used in the blank samples without enzyme extract.

Browning evaluations (BI)
BI of dried sweet potatoes was determined by Sarpong et al. [16], and the absorbance was monitored at 420 nm. The BI values were calculated by equations (11) and (12) as described by Ruangchakpet & Sajjaanantakul et al. [24].

Analyses of texture profiles (TPA)
The TPA of sweet potatoes was quantified by a texture analyzer (TA-XT2 of Stable MicroSystems, Ltd., Godalming, United Kingdom) equipped with a P/2N piercing plunger. The piston's speed and the stab force depth on the sample surface were 0.5 mm.sec -1 and 5.0 mm, respectively. The measurements were repeated four times for each sample. Hardness and resilience were assessed using Micro Stable software (Stable Micro Systems, Surrey, England).

Statistical study
Data are presented as mean ± SD and were checked by the Origin Pro 9.2 program. Two-way ANOVA and Tukey's test was used to compare the obtained data.

Mathematical modeling and kinetics of drying
The moisture level of sweet potato slices changed over time as a function of the experimental parameters, as shown in Figs. 1 and 2. As per the results obtained, it was increasing ultrasonic intensity and drying temperatures reduced the duration of drying and moisture level. During the ultrasound-assisted drying (USAD) of sweet potato slices, the pretreated ultrasound samples lost more moisture than the control samples due to texture degradation and the lack of a firm layer. HAD of sweet potato slices took 140 min at 70 • C using ultrasound frequencies (USF) of 20 and 40 kHz, while IRD took 50 min using USF of 20 and 40 kHz. The drying time of sweet potato slices were condensed to 150 min in HAD while 50 min in infrared drying (IRD) by applying 30 min of ultrasonic pretreatment. This demonstrates that longer drying durations were needed when the sweet potato slices were not pretreated with ultrasound. It has been reported that ultrasonic waves create the cavitation phenomena, causing dramatic fluctuation (expansions/ contractions) in the texture of the material and, as a result, transform the product's structure into a sponge-like tissue, allowing moisture to be removed from the product more quickly [5].
As illustrated in Figs. 1 and 2, the MC of sweet potato slices declined with elevation in temperature. The initial MC was high during the drying process, indicating a higher moisture content. This change is due to the material's internal water diffusion to the material's surface, and the MC gradually decreases with a decreasing gradient over time. The drying time was longer than at higher temperatures in the lower temperature range. Drying times reduce as the temperature rises because of the high  thermal gradient, giving a greater evaporation rate. The obtained outcomes are in line with previous reports conducted for drying numerous vegetables and fruits, e.g., sweet potatoes [26], apricots [27], almonds [8], and pineapple [28], using hot-air and infrared dryers with ultrasound pretreatment. Figs. 3 and 4 demonstrate the drying rate curves for pretreated ultrasonic samples dried in HAD and IR drying. It was discovered that US-treated samples had the highest drying rate throughout the drying period. Similar findings were made by Nowacka et al. [29] in dried ultrasonic pretreated apples. Initially, US pretreated samples dried in HAD and IRD (at US-40 kHz and 80 • C) had the highest drying rate, i.e., 13.05 and 16. 69 Kg/Kg dm, respectively. Hence, the US increased the moisture removal rate during convective drying. Both the control and US pretreated samples only showed the falling-rate period using HAD, whereas the samples dried with IR showed both the regular and falling-rate periods (Figs. 3 & 4). This stage ended after about 20-30 min of drying. When the samples were ultrasonically pretreated before convective drying, they may have taken up water. Water on the sample surface was loosely bound, causing the IR drying constant rate. The falling rate of the samples indicates that the drying approach is regulated by diffusion. Comparable findings were reported for cocoyam [30], sweet potato [9], corn, and wheat [31]. Sweet potato drying models were evaluated using the experimental data. Tables 1 and 2 show the non-linear regression applied to the empirical data statistical modeling.
There were four metrics used to evaluate models: R 2 (coefficient of determination), reduced chi-square (χ2), residual sum of squares (RSS), and root mean square error (RMSE). The Hii model was selected for sweet potato drying characteristics as it presented uppermost R 2 and lowest RSS (χ 2 ) and RMSE. This model gave the most acceptable representation of sweet potato slices' drying characteristics in both HAD and IRD dryers for pretreated ultrasound samples (Tables 1 and 2). Fig. 5 compares experimental MR for various samples at a drying temperature of 60, 70, and 80 • C. This trend provides vital support to the models' ability to predict the drying characteristics of sweet potatoes. Hii et al. [32] and Fan et al. [33] described the Hii model's accuracy in drying cocoa and sweet potato slices, respectively, with higher R 2 , lower RSS, χ 2 , and RMSE. The Two-term equation was the second best-fitting model in HAD after the Hii model. Dash et al. [34] made a similar discovery regarding the predictive ability of a two-term exponential model for star fruit drying performance.
On the other hand, the Page model fits the drying characteristics of ultrasound pretreated and control samples in IRD drying better than any different model. The Page model successfully predicted fresh cashew drying in a fixed bed dryer and can relate to the current study [35,36]. Whereas the other tested models were found unsuitable and suitable, tabulated in supplementary Tables S2 and S3 just for comparison.

E a analysis
The activation energy, or Ea, is the smallest energy required to overcome an energy barrier and start a process. The E a values ranged between 17.60 and 29.86 kJ/mol with an R 2 value of 0.999-0.9809. (Table-3). The investigation results fall within 11.57-36.44 kJ/mol, as stated by Onwude et al. [3], and 17.40-33.94 kJ/mol reported by Doymaz [37]. The samples recorded at 20 kHz and 40 kHz had the highest and lowest E a in HAD and IRD dryers. The lower the E a , the less energy is needed to achieve the active state. It accelerated mass transfer and diffusion by causing the products to dry more quickly (higher moisture effective diffusivity). Other studies have found a similar pattern [16,27]. Pretreatments, material type, and temperature affect activation energy [38]. Table 3 shows the thermodynamic functions (enthalpy, Gibbs, and entropy), demonstrating the minimum energy necessary to remove a certain amount of water. The natural logarithm of the rate constant vs 1/ K is evident from the Arrhenius plots ( Fig. 6a & b), implying that the slope of the plots indicated the hot-air and infrared drying dynamics. The enthalpy shifts (ΔH) (stated as the discrepancy in the reactant and the activated complex energy) were calculated using Eq (4). According to Table 3, ΔH values determined for HAD and IRD drying were similar, ranging from 8.60 to 14.91 kJ/mol for HAD and 18.77 to 33.77 kJ/mol for IRD dryers, respectively. Almost similar values were noted in dried banana slices ranging from 9.49 to 13.59 kJ/mol [16]. According to Eq. (4), temperature fluctuation had minimal influence on ΔH because of the less quantity of the ideal gas constant (R). As a result, differences in ΔH values were mainly attributable to variations in ultrasonic pretreatments, resulting in a direct relationship between ΔH and US. After ultrasonic pretreatment of sweet potato slices, the specific enthalpy of the drying process increased. Both dryers had decreased enthalpy values at 20 kHz US frequency, demonstrating the influence of low-frequency US pretreatments. However, the enthalpy was reduced when the drying temperature and US frequency were increased to 80 • C and 60 kHz. Furthermore, the enthalpy values were positive, consistent with Rashid et al. [9] findings for drying sweet potatoes. This finding supports the existence of endothermic reactions and the required heat to influence physical and chemical changes in vacuum-dried apples [39].

Thermodynamics
Gibb's free energy is a system activity unit in the adsorption practice that aids in the insight of the thermodynamic handling factors affecting the operations [39]. The ΔG values for HAD and IRD dryers ranged from 185.07 to 196.54 and 183.84 to 194.23 kJ/mol, respectively, indicating non-spontaneous reactions. The proximity of these outcomes suggests the rise in the system's total energy as the reagents approached. Gibb's free energy rose when the temperature increased from 60 • C to 80 • C in ultrasonic pretreated and control samples in both dryers. Drying is an endergonic reaction because it entails heat from the mediums, as shown by the positive Gibbs free energy values. One advantage of ultrasoundassisted treatments is that less heat is required to dehydrate the body's moisture. Nadi et al. [39] observed similar findings during the vacuum drying of apples.
Entropy is commonly expressed as the degree of disordering. This context is the level of disordering between moisture and the product. Entropy change (ΔS) for deterioration in sweet potato slices was used to calculate the disorder change ( Table 3). The modulus entropy of the sonicated samples decreased in the current study. Similarly, a decrease in entropy has also been reported in vacuum dried apples [39].
On the other hand, this phenomenon indicates that cell disruption confers an exemplary structure of the food material, implying that moisture can travel and generate zones of elevated moisture level. This reorganization may enhance the moisture gradients and the diffusion. As a result, the product dehydrates more easily. Additionally, rising temperature for drying augmented the entropy of the process. Similar effects were noticed by Rashid et al. [9] in this temperature range.

Consumption of specific energy
The quantity of SEC based on several treatments (ultrasonic pretreatments and varied temperatures) was used to compute by Eq. (19), and the results are tabulated in Table 3. The rise in temperature from 60 to 80 • C in HAD and IRD drying reduces the energy consumption in all ultrasonic pretreatments. Increasing the drying temperature of agricultural products reduces energy consumption [8]. There's also a clear correlation between increasing ultrasonic frequency and decreasing power consumption, as shown in Table 3. This is because ultrasonic waves do not form complex surface layers, which allows them to remove moisture from product tissues more quickly. Consequently, the drying process takes place in a short time and consumes less energy. The highest energy (14.08 kW/kg) was utilized at 60 • C, and the control treatment (CTL1) was in HAD.
Meanwhile, the lowest energy (7.04 kW/kg) was noted in the US-HAD6 treatment at 80 • C with a US frequency of 40 kHz. However, the control treatment consumed the most energy, i.e., CR1 (7.04) with infrared drying, and the lowest was noted in US-IR6 (2.93) at 40 kHz and a temperature of 80 • C. Santos et al. [40] also observed that as compared to Control, the energy was reduced by up to 53 % (Ethanol), 62 % (Ethanol + the US), and 41 % (Water + the US) for the ultrasound pretreatments used. These results confirmed the drying time and rate, as shown in Figs. 1 and 2.

Total phenolic compounds (TPC)
Table 4 depicts the differences in phytochemical components of dried sweet potato slices caused by various HAD and IRD drying combination techniques at three different temperatures (60, 70, and 80 • C) and US frequencies (20,40, and 60 kHz). After drying, TPC, TFC, and Vit-C levels significantly increased. In HAD, TPCs of fresh, HAD, and IRD samples ranged from 3.834 to 11.267 and 1.91,7.85-18.95 mg.GAE. 100.g − 1 dm, respectively. TPC content of dried samples was significantly (p < 0.05) elevated compared to fresh sweet potato samples. Likewise, Yang et al. [41] found a significant increase in the TPC of dried sweet potatoes. The highest TPC values i.e.,11.267 and 18.95 267 mg.GAE. 100.g-1 dm were obtained in the samples dried by US-HAD5 and US-IR5 (70 • C and 40 kHz USF), respectively, which could be ascribed to the homogeneous distribution and redistribution of temperature during the drying process. Similarly, ANOVA revealed a significant difference in TPC values among US-HAD, US-IRD, and control samples (p < 0.05). Due to the high temperature, TPC significantly decreased when phytochemical components were lost at 80 • C and USF 60 kHz. Thus, the lowest TPC values, i.e., 3.834 and 7.85 mg.GAE. 100.g − 1 dm were obtained in US-HAD9, and US-IR9 dried samples. This could be due to increased polyphenol breakdown from browning and oxidation reactions caused by prolonged exposure to hot air and infrared radiation. According to An et al. [42], electromagnetic radiation produces high and rapid heat, resulting in significant thermal degradation of phenolic compounds. Consequently, the most efficient temperature was 70 • C

Total flavonoid compounds (TFC)
TFCs increased significantly in all combined methods for US-HAD and US-IRD dried sweet potato slices ( Table 4). TFC of fresh sweet potatoes was 1.07. (mg.GAE. 100.g − 1 dm). TFCs of US-HAD and US-IR dried samples ranged from 2.76 to 14.19 and 7.27-10.15 mg.GAE. 100.g − 1 dm, respectively, was higher than fresh samples (1.07 mg.GAE. 100.g-1 dm). TFC values were found. The US-HAD5 and US-IR5 samples dried at 70 • C (USF 40 kHz) presented the highest TFC values, 14.19 and 10.15 mg.GAE. 100.g − 1 dm, respectively. The increase in flavonoid content caused by US pretreatments may be attributed to acoustic cavitation, retaining the mechanical stress on the samples, and allowing discharge of additional flavonoid compounds [43]. The results agreed with Ren et al. [6], who discovered that ultrasonic treatment of onions increased TFC levels. The intermittent combination of US-HAD and US-IRD would cause cellular compounds to break down, making flavonoids easily accessible during extraction [42]. Both IRD and HAD had a promising effect on TFC when samples were dried at 80 • C (p < 0.05). Flavonoid retention was increased due to the shorter drying times, temperature effect, and less heating used in US-HAD9 and US-IR9. Similarly, results were obtained when Chinese ginger was microwaved intermittently and convectively dried [42].

Vitamin C
The effect of the drying process and US treatments on the Vit-C content of dried sweet potato slices is shown in Table 4. All the dried samples had a significant (p < 0.05) reduction in vitamin C (ranging from 7.01 to 16.13 mg 100 g − 1 dm) except fresh sample. The HAD (US-HAD5) and IRD (US-IR5) dried samples (70 • C and USF 40 kHz) maintained the highest vit-C values (16.13 and 12.20 mg 100 -1 g dm, respectively). Ascorbic acid is well known for being relatively vulnerable to oxygen, light, and heat [44]. The Vit-C content in the US pretreated samples was comparable with the control sample (p < 0.05), indicating that US pretreatment had no beneficial effect on vit-C content due to its instability during processing of food material and susceptibility to decomposition by the influence of influence various factors including heat and light. Ercan & Soysal [45] found no significant difference in Vit-C content in the US treated samples and control. However, US-treated samples at 80 • C and 60 kHz frequency dried using both dryers retained significantly more vitamin C than those not treated. The reason could be that the cell wall was disrupted because of the alternative compression and expansion force on the samples because of ultrasound application [46]. The concentration of ascorbic acid in IRD samples was also significantly lesser than in HAD samples. Due to the inconsistency of the heating temperature during IRD drying speeds up the Vit-C oxidation, resulting in a frequent loss in it. As a result of the oxidation caused by heating, the infrared-drying temperatures negatively affected ascorbic acid retention. By combining US-treatment with IRD drying, Baeghbali et al. [15] noticed that infrared heating damaged more Vit-C in the apple slices than other procedures.

Total carotenoids contents (TCC)
The TCC of samples following various treatments and drying processes are presented in Table 4. Based on the findings, the US pretreatments with both used techniques have a significant impact on the deterioration of TCC. The retention of carotenoids in the current investigation altered from 0.91 to 6.48 mg 100.g − 1 dm in control and other samples processed with ultrasonic in both drying methods. Samples treated at 40 kHz by the US had higher total carotenoid content than untreated samples. This phenomenon might be associated with electroporation and sonication phenomena, resulting in more polar carotenoids (e.g., xanthophylls) leaking into the water treatment [47].
Furthermore, the degradation level of TCC was found to be dependent on the water activity of the samples, ranging between 0.5 and 0.6 mg 100.g − 1 dm. Sonication caused partial water evaporation, particularly on the samples. Hence, the moisture content of such pre-dried samples was reduced. Increased retention of sweet potatoes pretreated Means followed by different letters within columns are significantly different (n.s., *, ***, nonsignificant, or significant at p < 0.05, or p < 0.001, respectively).
before drying in the US may also be related to improved extractability of the carotenoids extraction from shattered chromoplasts [48] that changes the cell wall configuration [43]. The statistical assessment revealed that the type of US had the most significant effect on the variability of carotene retention.

β-carotene
Compared to fresh samples, the concentration of β-carotene increased after US pretreatment in all treatments, including controls (Table 4). At 80 • C, the significantly highest β-carotene concentration was found in US-HAD6 (5.19 µ. g − 1 dm), followed by US-HAD3 (4.38). A similar trend was seen in infrared drying, with the highest β-carotene concentration (3.92) obtained in US-IR6 at 80 • C. Wang et al. [49] discovered similar results for carrot samples, with carotene content increasing by 5.46 %, 13.07 %, and 17.62 % with ultrasonic treatment at 20 kHz. It was discovered that a shorter drying period at 80 • C preserved more carotenoids.
Additionally, it was found that the shorter drying period lowers the carotenoid loss in various varieties of sweet potatoes [50]. According to Wang et al. [49], β-carotene was discovered in the chromoplasts of carrot slices and was stabilized by lipoproteins. As a result, the increased β-carotene concentration was more than likely caused by ultrasonic extraction. The use of US on the sweet potato samples may have harmed the cellular structure, which was necessary for extracting carotene from the sample tissue during measurement. Table 5  The US-treated samples had the most significant antioxidant activities in terms of the ABTS, DPPH, FRAP, and reducing power analysis, as shown in Table 5. When the antioxidant capacities of the four different assays were compared, it was discovered that the pretreated sweet potato samples followed a pattern: US-HAD1 > US-HAD4 > US-IR1 > US-IR4. The samples' higher antioxidant activity (ABTS, DPPH, FRAP, and reducing power assays) might be attributable to the US-treated sample's shorter drying time than the control (as indicated previously), as the shorter drying period holds on a higher level of the antioxidant characteristics. These findings corroborate with An et al. [42], who described that food products must be dried for a shorter period and at a lower temperature (60 • C) to retain their antioxidant properties. According to Yilmaz et al. [52], numerous agricultural products that undergo the Maillard reaction after drying generally exhibit potent antioxidant activity due to a chain-breaking mechanism [53]. Our findings suggest enhanced antioxidant activities are often attributed to elevated TPC levels. As a result, the antioxidant content of the pretreated dried sweet potato samples is comparable to the TPC results. Ren et al. [6]  Effect of US pretreatments with HAD and IRD drying methods on enzyme inactivation of sweet potatoes. demonstrated that the higher antioxidant activity observed may improve antioxidant element extraction efficiency after drying. Wang et al. [38] found that US pretreatment improved antioxidant capacity in dried cherry tomatoes. Udomkun et al. [54] discovered that samples with intense browning had significantly higher antioxidant capacity than untreated samples after osmotic dehydration pretreatment of papaya.

Polyphenol Oxidase (PPO)
The PPO activity of dried samples was greater than that of fresh ones (Table 6). Generally, PPO activity is increased when the membranebound enzymes are tended to release. A similar effect was noted during the high-pressure processing of strawberries [55]. Regarding the impact of drying process factors on PPO activity, neither drying temperature nor air velocity demonstrated a significant trend. However, when the US frequency was enhanced, the PPO activity in sweet potatoes was reduced.
The procedure carried out at a US frequency of 60 kHz in both dryers at 80 • C resulted in the greatest drop-in PPO activity. With ultrasound treatment, more significant PPO activity was found at 70 • C in US-HAD5 and US-IR5 at 70 • C. Ultrasound of golden apple juice at 40 and 60 • C had a similar effect on PPO activity but a lower rate of inactivation (20 %) [56]. The disparity implies that the PPO of each apple type is susceptible to ultrasound treatment in various ways. In the high-pressure processing of plums and strawberries, a similar amount of inactivation was reported [55]. Because of the similarities in activation levels, the acoustic sonic technique is an appealing choice for PPO inactivation because it is less expensive than the high-pressure process. The activity of PPO in dried sweet potatoes showed a significant variation, supporting the impact of the acoustic treatment. When ultrasound was used, temperature changes did not affect PPO activity. Thermal processing, however, showed that as the temperature increased, the enzyme's activity dropped off significantly.

Peroxidase (POD)
Compared to control and fresh samples, sweet potato POD activity increased following ultrasound pretreatment (Table 6). An increase in POD activity could result from hydrogen peroxide (H2O2) generation via ultrasonic treatment since this enzyme is triggered by the rise in H 2 O 2 degrees, which is the enzyme's primary substrate. The present findings contrast with Jang & Moon [57], who discovered a slight decrease in the Peroxidase activity of Fuji apple samples treated with 40 kHz ultrasound. This disparity could be attributed to the different types of plant species compared to the current study and the different ultrasonic frequencies used in their research. PPOs are highly resistant to heat processing, significantly below 80 • C [58]. The results reported herein verify this and suggest that the energy delivered to the samples by ultrasonic treatment may decrease the endurable temperature of PODs.

Enzymatic and non-enzymatic browning
According to Table 6, there was a significant increase in enzymatic browning value at lower US-frequency (20 kHz), where a significantly (p < 0.05) high enzymatic browning value was identified in hot-air (US-HAD1, 0.69) and infrared (US-IR1, 4.38) drying. The decrease in enzymatic browning in samples was noted with the increase in US frequencies and temperature in both dryers. This loss may be explained by the extended drying period at high temperatures (70 and 80 • C) and the enzymatic browning response throughout the sweet potato drying process since temperature and drying time are essential variables in color degradation. In all US-pretreated samples, a comparable effect of temperature on the lowering of enzymatic levels was seen. POD and PPO enzymes cause enzymatic browning pigmentation as they affiliate with endogenic phenolic chemicals while drying the fruit samples [54].
Nevertheless, the contradictory effect of temperature and US frequency on the maintenance of non-enzymatic browning during the drying process was observed. Higher temperatures with lower US frequencies were seen to retain the non-enzymatic value compared to control samples, i.e., US-HAD (0.46 %) and US-IR3 (4.14 %) obtained the highest values at 20 kHz and 80 • C. This could cause the drying chamber's higher moisture level, diluting the released reactant species and delaying the chemical activities in damaged cells. Higher nonenzymatic browning values are most likely caused by brown polymers produced by the Maillard reaction. They rise at high internal temperatures of drying materials due to a drop in US frequencies. Udomkun et al. [54] proposed that phenolic acid and Vit-C oxidation might form complex brown pigments. Again, lower US frequencies encouraged nonenzymatic coloration as high values were obtained. This could be because the ultrasonic waves' massive disruption of cell integrity caused more reactant species to be released, resulting in further browning.

Browning index (BI)
The browning index of sweet potatoes was affected by various pretreatments and drying procedures (Table 4). Compared to HAD, IRD had a higher browning index. This could be associated with the Maillard reaction and the oxidation of Vit-C. In both dryers, ultrasound significantly reduced the BI values compared to the controls. Ultrasound pretreatment samples with lower BI values show that ultrasound slowed down BI production. Cernîşev [59] found the BI rate rose from 0.585 to 0.684 when dried tomato quarters were heated to various temperatures.

Surface color
The influence of temperature on the total color difference (ΔE) in US pretreated, HAD, and IRD dried samples is tabulated in Supplementary Table-S4. The 20 and 60 kHz ultrasound frequencies in the US-treated samples were significantly different in ΔE in US-IR1 at 80 • C (30.28), followed by US-IR7 at 60 • C (29.76). Compared to US pretreated samples dried in HAD, the IR drying technique with US pretreatments showed the higher ΔE values. Even though the drying period was shorter in IRD drying (Fig. 2), the drying temperature was high enough to intensify the non-enzymatic browning reaction, resulting in color alterations. When IR drying with US pretreatment was compared to HAD drying, the speed and intensity of ΔE were substantially faster. Due to the uneven nature of IR heating, the samples may burn during the procedure, and subsequently, color variation was examined more thoroughly than in the other samples. Rashid et al. [9] investigated the effect of US and IRD heating on sweet potato drying. They found that higher IR heating instigates high internal pressure, promoting color changes in the sample. Supplementary Table-S4 shows that a rise in the drying temperature induces higher ΔE values in all pretreated ultrasonic samples in HAD drying. Chemical reactions involving colorants, e.g., the Millard reaction and the synthesis of melanoidins during the drying, could explain the variation in the ΔE at various temperature ranges. These results are comparable with Bromberger et al. [60]. The rise in US frequencies (20-60 kHz) and IRD temperature range lead to a decline in ΔE. However, lower temperature influences the nutrients characteristics by lengthening the drying time, increasing ΔE. As a result, ultrasound does not affect ΔE by leaving chemical reactions unaltered; instead, it has a more physical function, and Cell wall disintegration happens in the samples' interior tissue. These findings are consistent with Bromberger et al. [60] and Shahidi et al. [10]. According to a previous study, dried products with a lower ΔE value are of higher quality [14]. Thus, despite being non-destructive, ultrasonic waves damaged the cell walls of tissues and caused the samples to dry faster when exposed to different temperature ranges. As a result, in various experimental treatments, the ΔE value was reduced, demonstrating the improved quality of the samples. According to statistical analysis (T × F interaction), significant differences (p < 0.05) have been found (Supplementary Table- 60,70 and 80 • C) and IRD samples were found significantly (p < 0.05) lower (or nearly similar in range) than fresh and control samples dried in HAD. According to Wiktor et al. [7], the h • values remained unchanged or higher for ultrasonic pretreated (by immersion) dried apple samples compared to the untreated sample. Kahraman et al. [61] reported that hue angle was significantly within different drying methods of apple slices, which supports our findings. The quantitative feature of colorfulness, known as chroma (C*), measures the degree of hue difference compared to a grey color of the same light. The results showed that the changes in C* were greater between the two drying methods for orange sweet potatoes. The samples treated in US-HAD and US-IRD were significantly lower than fresh samples. The US pretreated samples lost soluble solids (including color pigments), accounting for the lower chroma values. C* values were higher in the samples subjected to two drying techniques. The samples treated with US-HAD and US-IRD were much lower in C* values than fresh samples (Supplementary Table- Although the US pretreated and control samples were found nonsignificant, however, the C* values of the US pretreated samples were higher in both drying techniques than the control samples, which might be related to the influence of ultrasound cavitation, revealing the yellow pigment attached to the intracellular structure of the samples. Moreover, higher water loss leads to solid gain, resulting in color saturation. Costa et al. [62] demonstrated that the sonicated pineapple juice had a higher C* value than non-sonicated juice samples. samples (p > 0.05). Samples from both dryers significantly differed from the fresh samples (p > 0.05). Since ultrasonic pretreatment significantly improved the yellow appearance of the sweet potato, identifying the presence of β-carotene [63]. Except for a few, significant alteration in the whiteness index (WI) was observed amongst the samples. The WI of the US-HAD and US-IRD samples varied significantly (p < 0.05) compared with control and fresh samples. The higher WI values were noted in US-HAD5 (70.50) and US-IR5 (68.44) at 70oC and 40 kHz US frequency. Compared to the UStreated, WI values were lower in control samples, that might be related to the inactivation of peroxidase and polyphenol oxidase, which causes browning and thus preserves sample WI [64].

Texture profile analysis (TPA)
The hardness and resilience of US-pretreated samples did not vary significantly (p > 0.05) compared to untreated samples in hot-air and infrared drying (Table 7). According to Santacatalina et al. [65], increasing the amount of applied ultrasonic power led to a low but nonsignificant (p < 0.05) decline in sample hardness. Brncic et al. [66] found that increasing the US intensity from 50 % to 100 % amplitude increased the dried apple's hardness. Zhang et al. [67] discovered the same phenomenon with dried strawberry chips treated with the US having significantly higher hardness than control samples. Song et al. [68] linked the elevated hardness levels of the US pretreated samples with declined moisture content. The higher resilience in the matrix may account for the increased resilience in the US-pretreated samples. The findings revealed that sweet potato slices subjected to HAD drying, including control and US pretreated samples, had better textural qualities than infrared drying.

Conclusion
The current study comprised an innovative approach of drying sweet potatoes using ultrasound pretreatments with two drying methods since a synergistic application of these methods with a comprehensive set of experiments and fifteen predicting kinetic models were not conducted so far. Control samples dried slower and at a higher rate than ultrasoundtreated samples in both HAD and IR methods. The Hii model was the most accurate for predicting sweet potato drying for HAD and IRD drying. US frequencies, such as 20 and 40 kHz (US-HAD/IR 1-6) at 60 and 70 • C, were found to be appropriate, whereas 60 kHz (US-HAD/IR 7-9) US frequency (at 80 • C) resulted in poor quality attributes in terms of preservation and high hardness of the samples where specific pretreatment settings were insufficient. This study indicates that 20 min is a critical duration during which the drying rate in all ultrasonic treatments tends to decrease when compared to control samples. Sweet potato slices at 40 kHz US frequency in HAD demonstrated improved textural properties than IRD dried samples. Finally, this study showed that ultrasonic pretreatments combined with hot-air and infrared drying could result in an energy-efficient method with high-quality retention in dried sweet potatoes.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. s., *, ***, nonsignificant, or significant at p < 0.05, or p < 0.001, respectively).