American Ginseng Slice Drying and Optimization: Effect of Different Drying Methods on Drying Kinetics and Physicochemical Properties

Abstract

This study compared four drying methods, namely hot-air and vacuum combined drying (HAVCD), hot-air drying (HAD), vacuum drying (VD), and vacuum freeze-drying (VFD), with the aim to select a high-quality and efficient drying method to dry American ginseng slices. Drying kinetics and various physicochemical properties, such as color, rehydration ratio, total ginsenoside content, total ginsenoside residual rate, hardness, and microstructure were explored. An improved multi-indicator test formula method was used to score the experimental results quantitatively. HAVCD resulted in comprehensive advantages, with the highest product score for the dried American ginseng slices. Therefore, HAVCD was subjected to a response surface methodology (RSM) analysis and process optimization. The interaction of hot-air temperature and vacuum temperature on the comprehensive score of American ginseng drying was highly significant (p < 0.01). The optimized and validated process parameters obtained were a hot-air temperature of 57 °C, a vacuum temperature of 54 °C, and a moisture content at the conversion point of 39% (wet basis). Under these conditions, the best drying time was 170 min; the total ginsenoside content was 2.3 mg/100 g; the L* value was 91.68; the rehydration ratio was 3.10; and the comprehensive score was 96.77. The findings of this work indicated HAVCD as a promising drying method for American ginseng slices, considering the drying time and dried-product quality attributes.

1. Introduction

American ginseng (Panax quinquefolium L.) is a perennial herb in the Ginseng genus of the Araliaceae family, which is renowned for its health-promoting functions, such as blood nourishment, fluid production, physique strengthening, mind calming, and cognitive enhancement [1]. Being a very popular medicinal and edible herb, it has an important economic value [2], resulting in its use in health products in 35 countries globally. Ginsenoside is its primary active component, which is known to enhance immunity [3] and exhibit anticancer, anti-aging, and anti-fatigue activities [4], key points in its popularity [4]. Fresh American ginseng is high in moisture content (more than 70%, wet basis (w.b.)), which favors the growth of microorganisms. Furthermore, the high moisture content can easily enhance enzymatic and non-enzymatic reactions, resulting in its rapid deterioration and thus reducing its medicinal and commercial value [5,6,7]. Drying is a common preservation technology for maintaining the quality of agri-food products. Drying American ginseng can extend its shelf-life by reducing the moisture content to a low level (usually 10% w.b.), which prevents the growth of microorganisms and minimizes the costs of packaging, transportation, and storage [6,8]. Additionally, drying also affects the quality of American ginseng products, including color change, degradation of active ingredients, and rehydration ratio, which ultimately reduces their medicinal and commercial value [5,9].

Lu et al. [10] investigated the impact of the drying durations of the drying stages and the slice thickness of American ginseng during a vacuum freeze-drying process using a univariate experimental approach. They used ginsenoside Rb1 content as an indicator to optimize the process parameters. Vacuum freeze-drying was found to produce high-quality products, but it is characterized by a long drying time process, low efficiency, and high operational costs [8].

Hot-air drying equipment has a wide range of adaptations, and a simple and easy operation, making it widely used to dry American ginseng [7]. However, its high drying temperature in the presence of oxygen has a great influence on the quality of the medicinal components. In the case of American ginseng, hot-air drying causes ginsenoside degradation, color deterioration, shrinkage, and case hardening (which impede internal moisture migration) [6,7]. Du et al. [11] found that the color of American ginseng roots dried at 70 °C was dark, while the total ginsenoside content decreased with the increase in drying temperature.

Wang and Xiao et al. [6,12] investigated thin-layer air impingement drying (a type of hot-air drying) under different drying temperatures (35, 40, 45, 50, 55, 60, and 65 °C), air velocities (3, 6, 9, and 12 m/s), and sample thicknesses (1, 2, 3, and 4 mm) to analyze color, ginsenoside content, rehydration ratio, and microstructure. Their results indicated that the drying time was substantially shortened by increasing drying temperature and that both drying temperature and sample thickness had significant effects on the change in color, whereas air velocity did not have any significant effect. It was also observed that ginsenosides decreased with increasing drying temperature. Taking into account the drying time and dried sample quality, it was proposed that a drying temperature at 45 °C with a sample thickness of 2 mm should be used for thin-layer air impingement drying of American ginseng slices.

Although the above study effectively reduced the processing temperature, it was still in an aerobic heating environment throughout the process, which would still affect active ingredients like ginsenosides. Zhou et al. [13] found that the content of Rg1 (the main component of ginsenoside) in vacuum-dried samples was 17.1% higher than that in the hot-air-dried samples and suggested that this was due to the fact that vacuum drying avoids biochemical reactions, such as the oxidation of ginsenoside Rg1 during the drying process.

Kim et al. [14] compared the balanced low-pressure vacuum drying and hot-air drying of ginseng and found that a longer process time and a higher drying temperature of hot-air drying significantly reduced volatile bioactive phenolic compounds, which resulted in lowering its antioxidant activities. In contrast, balanced low-pressure vacuum drying improved the retention rate of bioactive components (such as acidic polysaccharides) to a level similar to that of freeze-drying. Furthermore, the liquid extracts from ginseng dried by balanced low-pressure vacuum drying increased the growth of human B and T cells as well as the secretion of both IL-6 and TNF-α. However, balanced low-pressure vacuum drying required a longer process time than that of hot-air drying. So, a single drying technology has disadvantages that are difficult to overcome.

Hot-air and vacuum combined drying (HAVCD) is an emerging technology developed in recent years [15]. It first uses hot-air drying to quickly remove a large amount of water in the material; then, it uses vacuum drying to create an anoxic environment to inhibit the occurrence of undesirable biochemical reactions, such as oxidative deterioration and browning reactions. Such a combination ultimately reduces the loss of active ingredients and improves the quality of dried products [16]. HAVCD can achieve the complementary advantages of different drying technologies, such as good product quality, high drying efficiency, and low energy consumption [15]. HAVCD has been successfully applied to dry a range of vegetables, such as lettuce [15]; celery [17]; as well as a number of Chinese herbal medicines, such as Astragalus membranaceus [18] and Scutellaria baicalensis [19]. However, to the best of our knowledge, there is no detailed report on the effects of HAVCD on American ginseng slices. Therefore, this study compared the effects of various drying methods, including HAVCD, on the properties of American ginseng slices.

Additionally, different drying technologies led to different microstructures of Chinese herbal medicines [8], which in turn, might affect the dissolution of active ingredients. In many countries, American ginseng is often used as a precious herbal medicine [2], which is soaked in water to dissolve its active ingredients for drinking. However, improper drying methods may affect the dissolution of ginsenosides like active ingredient, reducing its health or medicinal effects.

In order to select a high-quality and efficient drying method with optimized drying parameters, this study compared four drying methods, namely hot-air drying (HAD), vacuum drying (VD), HAVCD, and vacuum freeze-drying (VFD), based on various process parameters, including drying time, color, total ginsenoside content, total ginsenoside residual rate (which reflects the dissolution of ginsenosides), rehydration ratio, hardness, and microstructure. An improved multi-indicator test formula method was used to calculate and analyze comprehensive scores to select the optimal drying method. Then, the optimal drying method was analyzed through the response surface methodology (RSM), regression equations were established, the influence of factors on the experimental indicators was analyzed, and the processing parameters were optimized and verified. This study will provide a guidance for the application of new drying technologies to design a new processing equipment for drying American ginsengs with high quality, high efficiency, and low energy consumption.

2. Materials and Methods

2.1. Devices and Equipment

The main instruments used in this study included a BSA123S-CW one ten thousandth balance (Sartorius Scientific Instruments Co., Ltd., Beijing, China); KQ-300VDE ultrasonic cleaner (Kunshan Ultrasonic Instrument Co., Ltd., Kunshan, China); DR6000 UV-visible spectrophotometer (HACH Corporation, Loveland, CO, USA); TA.XT Plus texture analyzer (Stable Micro Systems Ltd., Godalming, Surrey GU7 1YL, UK); multifunctional grinder (Beijing Liren Technology Co., Ltd., Beijing, China); YS3060-type grating spectrophotometer (Guangdong SUNSHI Technology Co., Guangdong, China); FA-C electronic balance (accuracy 0.01 g) (Shanghai Youke Instrumentation Co., Shanghai, China); and TF-HFD-1A vacuum freeze-dryer (Shanghai Tianfeng Industrial Co., Ltd., Shanghai, China); Vega3-SBH Scanning Electron Microscope (TESCAN, Brno, Czech Republic).

The self-developed laboratory-scale HAVCD equipment was from the laboratory of Agricultural Products Drying Processing Technology and Equipment of Shaanxi University of Science and Technology. The same equipment was used to carry out HAD, VD, and HAVCD, which ensured consistency of the experimental conditions among the proposed three drying methods.

A scheme of the HAVCD equipment is shown in Figure 1. The equipment primarily consisted of a HAD system, a VD system, and a control system. Among them, the control system was designed based on the touch panel MT8071iE (WEINVIEW Co., Ltd., Shenzhen, China) and S7-200 PLC (Siemens (China) Co., Ltd., Beijing, China).

During hot-air drying, the electric vacuum isolation butterfly valve was open. Hot air was accelerated by the main fan and heated by a hot-air electric heating tube. Subsequently, hot air was evenly distributed via airflow to a uniform distribution chamber before being jetted into the combined drying chamber for material drying. Finally, the hot air was recycled back through the return air chamber to the main fan return air area for continuous circulation. The automatic control system utilized temperature sensors (model HC2A-IC 102 with ±0.1 °C temperature resolution, Michell Instruments (Shanghai) Co., Ltd, Shanghai, China) installed in the return air chamber to detect current temperature values according to the current temperature based on the proportional–integral–derivative (PID) function module in PLC to accurately control the power of the hot-air electric heating tube and the temperature in the combined drying chamber (temperature control accuracy of ±1 °C).

The VD stage mainly relied on hot water from the hot water tank to heat the material. Before VD, the hot water temperature was detected by the temperature sensor (model Pt100,with temperature resolution ± 0.1 °C, Meicon Automation Technology Co., Ltd., Hangzhou, China) and fed back to the control system, which accurately controlled the power of the electric heating tube and the hot water temperature based on the set hot water temperature with the PID function module in PLC (the temperature control accuracy was ±1° C). The VD system and the HAD system shared a combined drying chamber. During VD, the atmospheric air was isolated by closing the electric vacuum isolation butterfly valve and the pressure in the combined drying chamber was reduced to a preset value (vacuum sensor model PT500-503H with accuracy ±0.5 kPa, Meicon Automation Technology Co., Ltd., Hangzhou, China) by a vacuum pump. Subsequently, hot water was circulated from the hot water tank to the hollow material rack by means of a circulating water pump, which heated the metal trays placed on the hollow material rack and further heated and dried the material.

In the process of HAVCD, the equipment first entered the hot-air drying stage to quickly remove a large amount of free water in the material, enhancing the drying efficiency and minimizing the energy consumption. After the material moisture content reached the conversion point, the material did not need to be translocated, while the system was switched to the VD stage, isolating oxygen to avoid oxidation reaction, which effectively ensured the product quality. HAVCD proceeded until the moisture content of the material was reduced to a safe moisture content (less than 10%).

The above-mentioned fans, vacuum pumps, heating tubes, sensors, and valves were all connected to the control system, which were controlled and powered by the control system.

2.2. Materials

American ginseng (aged four years) was used for the experiment, which was sourced from an American ginseng plantation area in Hanzhong, Shaanxi Province (China). To mitigate the impact of material variability on the experimental outcomes, specimens of uniform size were selected for analysis (approximately 130–150 mm in length and 20–25 mm in diameter). The average initial moisture content was 72.0% (wet basis), as measured by VD at 70 °C for 24 h (AOAC, 1990) [6]. Prior to the experiment, the American ginseng was cleaned and drained, and the root head and lateral roots were excised and then cut into 2 mm segments [6], which were then placed in a sealed plastic bag (Figure 2). Mixing was performed to minimize the effect of individual differences in the ginseng roots on the experimental results. The samples were stored overnight (12 h) at 4 °C to equalize their moisture contents.

2.3. Experimental Procedure

2.3.1. Comparative Experiment

Approximately 200 g of American ginseng slices was obtained from each group. According to the results of Xiao’s study [6] on the optimal drying temperature of American ginseng slices, the drying temperature was chosen to be 45 °C; the air velocity for HAD was 3 m/s; and the vacuum level for VD was 10 kPa. HAVCD had the same HAD stage parameters and VD stage parameters as the aforementioned only-HAD or -VD parameters, respectively, with the moisture content at a conversion point of 30%. For the three drying methods mentioned above, the dehydrated samples were removed from the dryer to measure the moisture content; the drying was concluded once the moisture content fell below 10% [6,7,12,20].

VFD served as the control, which was acknowledged as the best drying method for dried product quality. For the VFD process, the material was frozen at −30 °C for 3 h, which was subjected to a primary drying process at −15 °C for 8 h and then to a secondary drying process at 30 °C for 5 h; a cold trap temperature of −45 °C and a vacuum of 1 Pa were used throughout the process [10]. Since VFD requires a continuous high-vacuum environment, no samples were taken during drying to detect moisture content in this study.

2.3.2. Response Surface Experimental Design

Optimization methods are often used to obtain high-quality dried products. Response surface methodology (RSM) is a collection of mathematical and statistical technologies for empirical modeling, which is a powerful method of process optimization and product improvement [15].

Following the Central Composite Design (CCD) guidelines, the variables examined were hot-air temperature (A), vacuum temperature (B), and moisture content at the conversion point (C). The evaluation indices included drying time (T), rehydration ratio (R), total ginsenoside content (G), color brightness (L*), and comprehensive score (Y). According to the results of Xiao’s study [6] on the optimal drying temperature of American ginseng slices, the drying temperature was chosen to be 45 °C. Referring to the study of HAVCD by Yuan and Yang [18,21], the moisture content at the conversion point was chosen to be 40%. And, the results of the one-factor test were the same as those in the above study. On this basis, a three-factor and five-level RSM experiment was designed with hot-air and vacuum temperatures of 45 °C, and the moisture content at the conversion point of 40% as the central point to analyze the influence of each factor on the indexes and to optimize the process parameters of the HAVCD of American ginseng slices by using its comprehensive score (Y). Subsequently, a validation test was conducted; the design of the test factors and levels is detailed in

Table 1. Coded values of corresponding actual values of independent variables.

Encodings	Considerations
	A (Hot-Air Temperature, °C)	B (Vacuum Temperature, °C)	C (Moisture Content at the Conversion Point, %)
−r	28	28	23
−1	35	35	30
0	45	45	40
+1	55	55	50
+r	62	62	57

.

2.3.3. Data Processing and Statistical Methods

Excel (Version 2019, Microsoft Redmond, WA, USA), Origin (Version 2022, OriginLab, Northampton, MA, USA), SPSS (Version 23.0, SPSS Inc., Chicago, IL, USA), and Design Expert (Version 8.0.6, Stat-Ease Minneapolis, MN, USA) software were used to statistically analyze the data, to make graphs and tables, to build models, and to test them. The indicators were normalized to a comprehensive score according to reference [22].

2.4. Drying Characteristic Analysis

2.4.1. Measurement of Moisture Content

Samples were weighted using an electronic balance (with an accuracy of 0.01 g) during drying. Drying was continued until the samples reached the desired final moisture content of 10.0% (w.b.). The product was cooled and heat-sealed in polyethylene (LDPE) bags. The experiments were performed in triplicate, and the means of the triplicate were used for drawing drying curves [6,7].

2.4.2. Dry Basis Moisture Content (M_t)

The dry basis moisture content of the samples was calculated from the wet basis moisture content (M_s) using the following formula:

$$M_t=\frac{M_S}{1-M_S}$$

(1)

2.4.3. Drying Rate

The drying rate (DR) was calculated according to Equation (2) [23].

$$DR=\frac{M_{t+{∆}_t}-M_t}{{∆}_t}$$

(2)

where M_t+Δt is the dry basis moisture content at moment t + Δt and Δt is the drying time of two adjacent moments.

2.4.4. Rehydration Ratio

Rehydration was carried out by immersing an approximately 5 g dried sample (random sampling) in distilled water at a constant temperature of 90 °C for 60 min. Then, the samples were removed from water, drained, and weighed [24]. Calculations were performed according to Equation (3) with three replications for each set of tests, and the results were averaged.

$$RR=\frac{m_2}{m_1}$$

(3)

where RR is rehydration ratio; m₁ is the mass (g) after rehydration; and m₂ is the mass (g) before rehydration.

2.5. Determination of Ginsenoside Content

A sample (random sampling, about 1 g) was taken in a 100 mL volumetric flask and sonicated for 30 min with a little bit of water. Water was then added until the total volume reached 100 mL. After centrifugation at 6000 rpm for 10 min, a 1 mL aliquot of the supernatant was used for column chromatography. For column chromatography, a 10 mL syringe served as the chromatographic column, containing 3 cm of Amberlite-XAD-2 macroporous resin with 1 cm of neutral alumina. The column was first washed with 25 mL of 70% ethanol and then with 25 mL of water; both eluates were discarded. After that, 1.0 mL of the sample solution was eluted first with 25 mL of water and the elute was discarded. Then, 25 mL of 70% ethanol was used to collect the elute in an evaporating dish, which was evaporated to dryness in a water bath at 60 °C; such a dried sample was subsequently used for the ginsenoside analysis. Vanillin in glacial acetic acid (0.2 mL, 5%) was added to the evaporating disc to dissolve the dried residues, and 0.8 mL of perchloric acid was mixed thoroughly, which was then transferred to a 5.0 mL stoppered centrifuge tube. After heating for 10 min at 60 °C and cooling, 5.0 mL of glacial acetic acid was added and vigorously shaken. A colorimetric analysis was conducted alongside the standard tube using a 1 cm quartz cuvette at a wavelength of 560 nm. A ginsenoside Re standard solution (2.0 mg/mL) was dispensed in 100 μL aliquots into an evaporating dish and dried using a water bath (below 60 °C) [25]. The total ginsenoside content was calculated using Equation (4).

$$x=\frac{A_{1}CV}{10000A_{2}m}$$

(4)

where x is the total ginsenoside content of the sample (g/100 g); A₁ is the absorbance of the measured solution; A₂ is the absorbance of the standard solution; C is the standard tube ginsenoside Re amount (μg); V is the volume of sample dilution (mL); and m is the sample mass (g).

2.6. Calculation of Total Ginsenoside Residual Ratio

The total ginsenoside residual rate (L) was defined as follows: Rp represented the total ginsenoside detection value (g/100 g), which was extracted and determined under the optimal dissolution condition (American ginseng powder) and referred to the total ginsenoside content in the sample; R_s was the total ginsenoside content (g/100 g) extracted and detected in the sample under different drying methods, which reflected the extractable content of total ginsenosides in the sliced samples. R_p values were always greater than Rs values, and the difference between R_p and R_s largely reflected the loss of total ginsenoside dissolution due to differences in the pore structure. This loss, when divided by R_p, gave the percentage of total ginsenoside loss, as demonstrated in Formula (5):

$$L=\frac{{R_p-R}_s}{R_p}$$

(5)

The ginsenoside residual rate (L) represented the effect of different drying methods on the degree dissolution of ginsenosides in American ginseng slices.

2.7. Measurement of Hardness

The hardness was determined using a TA.XT Plus texture analyzer, employing the TPA (texture profile analysis) mode with an A/CKB type probe. The probe descended to the sample surface at 2 mm/s and continued downward at 1 mm/s. Upon reaching 70% compression, the probe returned to the starting position. After a 3 s pause, the second phase commenced, replicating the same compression before returning to the zero point at 2 mm/s [20]. This procedure was replicated eight times for each sample (random sampling), and the final results were averaged.

2.8. Measurement of Color

A YS3060 grating spectrophotometer was employed to quantify the surface brightness L* (0 < L* < 100). For American ginseng slices, the higher the L* value, the whiter the color, indicating more favorable consumer perception [7]. Sampling for color measurement was performed at the point when the chrysanthemum pattern was in the center of the American ginseng slices. Three sample slices (random sampling) were taken for each drying method, and the results were averaged.

2.9. Microstructure Analysis

Scanning electron micrographs showed the microstructural features of the samples. The planar microstructures of the samples were observed at an accelerating voltage of 10 kV. The sample slices were gold-sprayed for 40 s before testing [26]. The micrographs were obtained at 200× magnification, and representative micrographs were selected to represent each sample.

2.10. Synthesized Assessment

A multi-indicator test formula method can objectively assign weights to the indicators without changing the order of the indicators and the optimal program, which evaluate the test effect in the form of a comprehensive score [27]. However, this method results in the indicator nature of large differences in the calculated weight coefficients being huge. Small extremum difference indicators’ weight coefficients are several times or even dozens of times higher than those of the large extremum difference indicators, which leads to the final score being too sensitive to the small extreme indicators. The coefficient of variation method can give appropriate weighting coefficients according to the size of the numerical extremes of different indicators. Therefore, this study applied the coefficient of variation method to improve this step.

The improved multi-indicator test formula method [22] was as follows: The optimal value, denoted as (xb), of each index was assigned a full score of 100 points. Subsequent values were then converted to scores, represented as (X_ij), with no score exceeding 100. Indices were classified as either positive (higher values indicated better quality) or negative (lower values indicated better quality).

Positive indicator score conversion formula:

$$X_{ij}=\frac{100x_{ij}}{x_b}$$

(6)

Negative indicator score conversion formula:

$$X_{ij}=\frac{100x_b}{x_{ij}}$$

(7)

where X_ij is the score of the jth indicator for the ith set of trials; x_ij is the measured value of the jth indicator for the ith set of tests; and x_b is the optimization of the measured value of each indicator.

The coefficient of variation method was used to objectively assign weights to each indicator:

$$W_i=\frac{V\left(x_j\right)}{\sum _{j=1}^{n}V\left(x_j\right)}$$

(8)

where W_i is the weighting coefficient for the ith indicator and V(x_j) is the coefficient of variation for the jth indicator.

The comprehensive score P_i was calculated for each group of trials:

$$P_i=\sum \left(X_{ij}\times W_i\right)$$

(9)

3. Results and Discussion

3.1. Characterization of Dried American Ginseng Slices Using Different Drying Methods

Figure 3 illustrates the drying curves of American ginseng slices. As observed in Figure 3a, under the test conditions, HAD showed the shortest drying time, requiring about 150 min; VD took a longer time (223 min); and HAVCD was between the two (180 min). The HAVCD curve almost overlapped with the pure HAD curve in the HAD stage, and after the conversion point for moisture content, it slowed down significantly. Yang et al. [21] also reported a trend during banana slice drying, which was attributed to the lower drying rate of VD, compared to HAD; the reason was that during the VD process, the circulating hot water was used to heat the hollow metal material rack and the metal material trays, from which the metal material trays mainly heated the surface of the material in the form of contact conduction, which was then conducted to the inside of the material. VD can reduce the boiling point of water by lowering the vacuum level through evacuation. However, with a continuous improvement in the drying degree of the material, the thermal conductivity and heat condition of the material continue to deteriorate. A similar phenomenon was found and explained in the effects of pulsed vacuum drying based on electronic panel contact heating in the study conducted by Xie et al. [28]. Meanwhile, the constant vacuum pressure in this study created a balance in the surface aqueous partial pressure and decreased the gradient of aqueous partial pressure; thus, the efficiency of dehydration was decreased [28]. In contrast, in HAD, because hot air is a low mass fluid, the hot air with a certain velocity could form a thin airflow boundary layer on the surface of the material, which could rapidly transfer heat to the material by convective heat transfer and quickly take away the evaporated moisture from the material, thus continuously maintaining a higher drying speed [6,12].

From Figure 3b, it was found that the drying rate of American ginseng slices under HAD, VD, and HAVCD decreased gradually with a decrease in moisture content, which indicated that the whole drying process occurred in a falling rate period, indicating no distinct constant-rate drying phase. Wang et al. [12] also reported the same conclusion in their study on the drying of American ginseng. The primary mechanism governing the drying was internal moisture diffusion. That is, most of the time, after the material was heated, the moisture needed to be diffused to the evaporating surface of the material from the inside and then evaporated, while the drying rate was controlled by the diffusion rate of the moisture inside the material. In HAD, the heat was conducted from the surface of the material to the material inside, and the moisture inside the material diffused to the evaporation surface of the material periphery to evaporate under the action of humidity difference and heat [29]. In VD, under the effects of a pressure difference, a humidity difference, and the internal moisture of the material being heated, the internal moisture of the material diffused to the evaporation surface of the material and then evaporated to the low-pressure air inside the vacuum box body and was pumped away by the vacuum pump [29]. The internal moisture diffusion mechanism of the HAVCD process was the same as that of only HAD or VD.

3.2. Comparative Study of Different Drying Methods for American Ginseng Slices

Table 2. Comparison of American ginseng slices dried with different methods.

Drying Methods	Drying Time (min)	L*	Total Ginsenoside Content (g/100 g)	Ginsenoside Residue Rate (%)	RR	Hardness (N)	Comprehensive Score
HAD	150 (100.00)	82.86 ± 0.24 b (86.87)	1.79 ± 0.07 c (70.20)	48.81 ± 0.14 b (67.34)	2.50 ± 0.07 c (81.43)	2507.58 ± 5.8 c (60.11)	83.25
VD	223 (67.26)	90.42 ± 0.28 a (94.80)	1.89 ± 0.06 c (74.12)	52.50 ± 0.19 a (62.61)	2.44 ± 0.09 c (79.48)	4171.97 ± 9.1 a (100.00)	75.31
HAVCD	180 (83.33)	92.67 ± 0.31 a (97.16)	2.12 ± 0.10 b (83.14)	42.19 ± 0.17 c (77.91)	2.80 ± 0.07 b (91.21)	3820.76 ± 11.4 b (91.58)	84.56
VFD	900 (16.67)	95.38 ± 0.29 a (100.0)	2.55 ± 0.08 a (100.00)	32.87 ± 0.15 d (100.00)	3.07 ± 0.10 a (100.00)	1265.3 ± 7.3 d (30.33)	41.17
Coefficient of variation	0.87	0.05	0.14	0.17	0.09	0.38
Weighting coefficients	0.51	0.03	0.08	0.1	0.05	0.22

Note: Different lowercase letters after data in the same column in the table indicate significant differences (p < 0.05).

presents the comprehensive scores for American ginseng slices dried with different methods. Drying time is a crucial metric for assessing the efficiency and energy consumption of drying methods. Under the test conditions, the VFD duration was the longest, being six times longer than that of HAD. Low efficiency and high energy consumption are significant barriers to the large-scale adoption of VFD [12]. VD time ranked second to VFD; however, the drying time for HAVCD was shorter by 43 min (19.3%), compared with VD alone.

For American ginseng slices, color is a primary factor influencing the consumer assessment of product quality. In terms of color values, VFD American ginseng slices exhibited the highest L* value (as shown in Figure 4a), and their color was also the whitest; HAVCD yielded better L* values than only HAD or VD (as shown in Figure 4b–d), which shows that the color of the HAVCD samples was superior to those of only HAD or VD. Such results might be due to the fact that the color of American ginseng slices is sensitive to oxidation reactions as well as drying time. Xiao et al. [6], in their study on air jet impingement drying of American ginseng, pointed out that prolonged drying time adversely affects the color index (including L* value). Compared with VD, HAVCD significantly reduced the drying time (19.3%), which therefore resulted in a better color appearance. Meanwhile, Jiang et al. [16] found that samples dried by vacuum pulsation drying had better brightness than those dried by hot air, which was thought to possibly be due to the fact that the vacuum condition during drying effectively avoided contact with oxygen, inhibiting the occurrence of an oxidation reaction. The HAVCD was under vacuum isolation from oxygen (44.4%), which reduced the oxidation reaction, and therefore, its color index was better than that of HAD.

Ginsenoside content is an important indicator of the quality of American ginseng [6,10]. Vacuum freeze-dried American ginseng exhibited the highest total ginsenoside content. Among the other tested drying methods, HAVCD of American ginseng had a higher total ginsenoside content than only HAD or VD (18.4% and 12.2% higher, respectively). This was attributable to the physicochemical properties of ginsenosides and the characteristics of the drying methods. Firstly, researchers found that ginsenosides are thermally unstable [7,30], and the total ginsenoside content of ginseng roots decreased from 7.0 to 4.5% when the drying temperature was increased from 45 to 60 °C. This finding was in good agreement with those reported by Xiao et al. [6] and Hwang et al. [31]. Therefore, under identical heating conditions, shorter drying durations were more conducive to ginsenoside retention. At the same time, Zhou et al. [13] found that Rg1 (the main component of ginsenoside) in the samples was better retained in the VD-isolated oxygen environment than in the HAD aerobic environment. HAVCD showed a shorter drying time by 19.3% than VD, significantly enhancing the ginsenoside content. Furthermore, although the drying time of HAD was 16.7% shorter than that of HAVCD, the proportion of HAVCD in the vacuum stage was as high as 44.4%. A long VD period resulted in an avoidance of oxygen, inhibiting the oxygen-demanding type of biochemical reaction and ginsenoside degradation [6,32]. Therefore, the total ginsenoside content in the HAVCD samples was higher than that in HAD samples.

The smaller the total ginsenoside residue rate, the higher the proportion of total ginsenoside that can be solubilized when dried American ginseng slices were soaked in water for drinking, which resulted in a better health effect. As shown in Table 2, different drying methods resulted in variations in the residual rates of total ginsenosides in American ginseng slices, and VFD resulted in the lowest total ginsenoside residue. Compared with HAD and VD, the total ginsenoside residue of HAVCD was smaller and reduced by 13.6% and 19.6%, respectively, indicating that the HAVCD samples had higher dissolution ratios of total ginsenosides when dried American ginseng slices were soaked in water for drinking.

The rehydration ratio of the VFD American ginseng slices was the highest, followed by that of HAVCD. The rehydration ratios of HAVCD surpassed those of HAD and VD by 12% and 14.8%, respectively. This phenomenon was basically similar to the experimental results of the ginsenoside residual rate. Previous studies have shown that different drying technologies resulted in the formation of different microstructures in the material [8,33], and the rehydration rate of dried products was largely influenced by the microporous structure of the dried products [6,20,34]. This was probably because, during the rehydration process, the water needs to enter the internal microstructure of the material. The smaller the destruction of the microstructure (shrinkage, closure, collapse, and so on) in the drying process, the more adequately the water entered the microporous channels of the material, the better the combination with the material in the rehydration process, and the higher the rehydration ratio. At the same time, the contact and dissolution of the active ingredients in American ginseng with water would be more adequate and more ginsenosides would be dissolved, reducing the ginsenoside residue rate. Therefore, it can be summarized that the ginsenoside residual rate and the rehydration rate were largely affected by the microstructure.

As shown in Table 2, VFD showed the best rehydration rate and ginsenoside residue rate because it formed a loose and porous microstructure (as shown in Figure 5a). The electron microscopy results showed that the pore distribution of vacuum freeze-dried American ginseng was more uniform, mainly due to the fact that the VFD process was completed in a freezing state; its tissue structure was less damaged; and the cells were mostly elliptical, with the largest pores, which was in agreement with the results of the study by Zhang et al. [35]. HAD had the shortest drying time, and the material was subjected to a strong convective heat transfer period during the drying process, forming a dense pore structure [7]. As shown in Figure 5c, the electron microscopy results showed that the cellular pore deformation of hot-air-dried American ginseng slices were large, the pore size was relatively small, and there were more pore adhesion closure areas [6]. Jiang et al. [8] concluded that, at the same drying temperature, a longer drying time resulted in an accumulation of more heat, leading to a collapse in the microporous matrix structure in the product, which caused more compact microstructures and impeded the reabsorption of water. The VD time was the longest, and the material was subjected to heat for the longest time at the same temperature, which was 48.7% and 23.9% longer, compared with the HAD and HAVCD times, respectively. Thus, the microstructure of the vacuum-dried samples was denser. For example, in Figure 5d, the electron microscopy results showed that vacuum-dried slices of American ginseng showed marked cell shrinkage, furrows caused by curling, and marked matrix collapse, which led to the rehydration rate and ginsenoside residue rate of VD being the worst among the four drying techniques. The rehydration rate and total ginsenoside residue rate of HAVCD were better than those of VD and HAD, which might be due to the fact that the drying time of HAVCD was in between those of HAD and VD, without the sharp shrinkage of rapid drying or the large amount of heat accumulation of prolonged drying, and thus formed a more loose and porous microstructure of the product. As shown in Figure 5b, the electron microscopy results showed that the microporosity of HAVCD was not as good as that of VFD, and cell wall collapse occurred in some areas but was better than in only HAD or VD.

The hardness of the vacuum-dried American ginseng slices were the highest, followed by HAVCD, while that of VFD American ginseng slices were the least because of their loose microstructure [35]. However, such products were easy to break and shatter during distribution and storage [13]. The highest hardness values for the vacuum-dried samples might be due to the longest VD time, resulting in a dense internal structure of their samples [8]. The hardness of the HAD sample was obviously lower than that of VD, which might be due to a rapid evaporation of water inside the material during HAD, which would produce an expansion effect, resulting in lower hardness [34]. The hardness of HAVCD was lower than VD but higher than HAD, which might be due to the joint effect of HAD and VD.

According to the comprehensive score of each index (Table 2), it was seen that the overall drying effect of the HAVCD technology was good, showing an improved color; an increased ginsenoside content; a decreased ginsenoside residual rate; an increased rehydration; and better pore structure. Moreover, the HAVCD greatly shortened the drying time, resulting in a lower energy consumption for drying. Therefore, we selected HAVCD as a high-quality and high-efficiency drying technology for American ginseng slices for further optimization study of its processing parameters.

3.3. Optimization of Hot-Air and Vacuum Combined Drying (HAVCD) Technology for American Ginseng Slices

3.3.1. Central-Composite Experimental Design and Results

The response surface comprised 20 test points, including 14 factorial points to investigate factor interactions and 6 center points to estimate the experimental error. The obtained results are shown in

Table 3. Experimental design and results for the response surface test.

Serial Number	A (°C)	B (°C)	C (%)	T Drying Time (Min)	RR	G Total Ginsenoside Content/(g/100 g)	L*	Y Comprehensive Score
1	35	35	30	315 (54.92)	2.82 (90.91)	1.70 (70.54)	92.95 (99.25)	64.25
2	55	35	30	282 (61.35)	2.83 (91.12)	1.74 (72.20)	91.00 (97.17)	68.33
3	35	55	30	252 (68.65)	3.06 (98.53)	1.22 (50.76)	90.96 (97.13)	65.63
4	55	55	30	177 (97.74)	3.11 (100.00)	2.19 (90.73)	89.71 (95.79)	95.47
5	35	35	50	402 (43.03)	3.02 (97.12)	1.89 (78.42)	92.00 (98.23)	60.82
6	55	35	50	381 (45.41)	2.92 (94.02)	1.62 (67.36)	91.01 (97.18)	58.08
7	35	55	50	270 (64.04)	2.78 (89.61)	1.71 (70.82)	90.62 (96.77)	69.24
8	55	55	50	205 (84.39)	2.77 (89.14)	2.36 (97.93)	90.09 (96.20)	89.70
9	28	45	40	284 (60.92)	3.00 (96.60)	1.62 (67.22)	92.24 (98.49)	66.86
10	62	45	40	200 (86.50)	2.95 (94.98)	2.18 (90.46)	90.55 (96.69)	88.79
11	45	28	40	336 (51.49)	3.04 (97.88)	1.83 (75.93)	92.52 (98.79)	64.73
12	45	62	40	173 (100.00)	3.09 (99.45)	2.05 (85.06)	90.82 (96.98)	94.77
13	45	45	23	220 (78.64)	2.97 (95.63)	1.82 (75.52)	92.68 (98.96)	79.44
14	45	45	27	285 (60.70)	2.80 (90.16)	2.10 (87.14)	91.99 (98.23)	73.03
15	45	45	40	205 (84.39)	2.91 (93.55)	2.20 (91.29)	92.49 (98.76)	87.84
16	45	45	40	210 (82.38)	2.92 (94.01)	1.95 (80.91)	93.65 (100.00)	83.26
17	45	45	40	227 (76.21)	2.97 (95.57)	2.33 (96.68)	92.77 (99.06)	85.33
18	45	45	40	232 (74.57)	2.83 (91.20)	2.41 (100.00)	92.52 (98.79)	85.20
19	45	45	40	245 (70.61)	2.90 (93.30)	2.02 (83.82)	92.29 (98.54)	77.64
20	45	45	40	229 (75.55)	2.88 (92.79)	1.99 (82.57)	92.48 (98.75)	79.91
Coefficient of variation				0.24	0.03	0.15	0.01
Weighting coefficients				0.55	0.08	0.34	0.03

. A multiple regression fitting analysis was performed on the data using Design-Expert 8.0.6 software and a quadratic model was chosen, considering hot-air temperature (A), vacuum temperature (B), and moisture content at the conversion point (C) as independent variables, while comprehensive score (Y) as a dependent variable. The following quadratic regression equation was established:

$$\mathrm{Y}=83.4+6.48\mathrm{A}+8.72\mathrm{B}-1.95\mathrm{C}+6.12\mathrm{A}\mathrm{B}-2.03\mathrm{A}\mathrm{C}+1.44\mathrm{B}\mathrm{C}-3.2{\mathrm{A}}^2-2.52{\mathrm{B}}^2-3.76{\mathrm{C}}^2$$

(10)

3.3.2. Significance Analysis of Regression Equations

The response surface experimental results are shown in Table 3. The test indexes were normalized into comprehensive scores based on an improved multi-indicator test formula method.

To evaluate the fitness of the model, the significance of the main effects was determined and factor interactions were identified. Regression equations were subjected to an analysis of variance (

Table 4. Variance analysis of comprehensive scores.

Source of Variance	Square Sum	Degrees of Freedom	Mean Square	F	p (p > F)
Model	2386.19	9	265.13	13.60	0.0002
A	573.80	1	573.80	29.44	0.0003
B	1038.39	1	1038.39	53.28	<0.0001
C	51.85	1	51.85	2.66	0.1339
AB	299.53	1	299.53	15.37	0.0029
AC	32.84	1	32.84	1.69	0.2234
BC	16.56	1	16.56	0.85	0.3783
A2	147.87	1	147.87	7.59	0.0203
B2	91.63	1	91.63	4.70	0.0553
C2	204.15	1	204.15	10.47	0.0089
Residuals	194.89	10	19.49
Lost proposal	123.07	5	24.61	1.71	0.2845
Pure terror	71.83	5	14.37
Total deviation	2581.08	19
Ratio	R2 = 0.9245; Adj R2 = 0.8565 Adeq Precision = 12.053

). The p-value of the model was 0.0002, indicating highly significance; the lack-of-fit term was not significant at the p < 0.05 level. R² was 0.8565, explaining the majority of variance in the response values. The F-test results indicate the following order of factor contributions: vacuum temperature > hot-air temperature > moisture content at the conversion point.

3.3.3. Response Surface Analysis

To investigate the interactions among the factors, one factor was held constant at the zero level and response surface and contour plots were constructed for AB, AC, and CB against the comprehensive score based on the regression model. From Figure 6a, combined with an analysis of variance (ANOVA), it can be seen that the interaction of hot-air temperature (A) and vacuum temperature (B) for the comprehensive score was highly significant (p < 0.01). As can be seen from Figure 6a, the comprehensive score (Y) increased with the increase in hot-air temperature (A) and vacuum temperature (B) at a certain moisture content at the conversion point (C). The maximum value in Figure 6a appeared when both hot-air temperature (A) and vacuum temperature (B) were 55 °C, with the maximum value of the comprehensive score at 98.9304. From Figure 6b,c, combined with ANOVA, it can be seen that the interaction of moisture content at the conversion point (C) with vacuum temperature (B) and hot-air temperature (A) was not significant (p > 0.05). Compared with Figure 6b,c, the response surface of Figure 6a was steeper, which also indicated that the interaction effect of hot-air temperature (A) and vacuum temperature (B) on the comprehensive score (Y) was more significant [32,36]. As seen from Figure 6b, under a certain hot-air temperature, the comprehensive score gradually increased with the increase in vacuum temperature (B). With the increase in moisture content at the conversion point (C), the comprehensive score showed a trend of first increasing and then decreasing, and the maximum value of the response surface was 89.9304. The response surface showed a trend of first increasing and then decreasing, and in the case of a vacuum temperature (B) of 55 °C and a moisture content at the conversion point (C) of 39.3%, the maximum value of the response surface was 89.6. In the case of a certain vacuum temperature (B), the comprehensive score (Y) increased with the increase in hot-air temperature (A), while the comprehensive score (Y) showed a tendency to increase first and then to decrease with the increase in moisture content at the conversion point (C) (Figure 6c). In the case of a hot-air temperature (A) of 55 °C and a moisture content at the conversion point (C) of 34.72%, the maximum value of the response surface was 87.7.

3.3.4. Optimization and Validation of HAVCD Process

The selected objective function and constraints are shown in Equation (11) and Equation (12), respectively. The regression Equation (10) was further analyzed and predicted using Design-Expert 8.0.6 software to obtain the following optimized process conditions: hot-air temperature of 57.13 °C; vacuum temperature of 54.11 °C; and moisture content at the conversion point of 39.73%. Additionally, the model predicted the comprehensive score of dried American ginseng slices under these conditions to be 99.24.

$$\mathrm{Y}\to {\mathrm{Y}}_{\mathrm{m}\mathrm{a}\mathrm{x}}$$

(11)

$$\left\{\begin{array}{c}28\le \mathrm{A}\le 62\\ 28\le \mathrm{B}\le 62\\ 23\le \mathrm{C}\le 57\end{array}\right.$$

(12)

The regression equation was further analyzed with Design-Expert 8.0.6 software to obtain the following optimized process conditions: hot-air temperature of 57.13 °C; vacuum temperature of 54.11 °C; and moisture content at the conversion point of 39.73%. Additionally, the model predicted the comprehensive score of dried American ginseng slices under these conditions to be 99.24.

In order to further verify the accuracy and validity of the regression equation, a validation test of the optimal HAVCD conditions was conducted. Considering the convenience of a practical operation, the validation test was conducted under the conditions of a hot-air temperature of 57 °C, a vacuum temperature of 54 °C, and a moisture content at the conversion point of 39%, which resulted in a drying time of 170 min, a total ginsenoside content of 2.3 mg/100 g, a color brightness of 91.68, a rehydration ratio of 3.10, and a comprehensive score of 96.77, with a relative error of 0.5%. The experimental value was basically close to the predicted value of the model, indicating that the model was reliable. Additionally, the values of drying time, total ginsenoside content, color brightness, and rehydration ratio for HAVCD were all better than those for a only-VD product with 45 °C as the processing temperature.

4. Conclusions

This study compared the drying of American ginseng slices through the HAVCD, HAD, VD, and VFD methods while considering drying time, color, rehydration, total ginsenoside content, total ginsenoside residual rate, hardness, and microstructure as the evaluation parameters. Although VFD (control method) produced the best-quality dried products, its longer drying time caused a higher energy consumption, restricting its use for commercial application. Among the other tested drying methods, HAVCD showed comprehensive advantages and the highest score for the dried products; in particular, the samples dried using this method had the highest total ginsenoside contents and the lowest total ginsenoside residue rates, resulting in better health effects when soaked in water for drinking. The CCD RSM of HAVCD indicated optimized and validated process parameters, showing the best evaluation parameters. Therefore, HAVCD is a processing technology suitable for American ginseng and has a good application prospect. This study is expected to provide guidance for the application of HAVCD for the drying of American ginseng as well as for the design of new HAVCD processing equipment.