Ultrasound Assisted Osmotic Dehydration as Pretreatment for Hot-air Drying of Carrot

Abstract

Ultrasound assisted osmotic dehydration (UOD) combined hot-air drying (AD) process was investigated in this study. The influences of operation parameters such as ultrasonic energy density and solution concentration on weight reduction ratio (WRR) were discussed. The effects of UOD on the following AD of carrot slices were also studied. The results showed that ultrasonic energy density and solution concentration have positive and significant effects on WRR. Osmosis dehydration (OD) pretreatment without ultrasound assistance before AD has negative effects on the total processing time and the effective moisture diffusivity. Yet when ultrasound is applied, the increase of ultrasonic energy density could shorten dehydration time of the following AD process and total processing time. UOD prior to AD has positive effects on improving carotenoid content of product and reducing process energy cost. So it's definitely concluded that UOD pretreatment is an effective and complementary method for traditional AD process.

Introduction

At the present time, consumption of dehydrated food has experienced a noticeable increase in its demand on the market. Conventional dehydration methods based on hot-air drying (AD) are widely used in agricultural industry, but they are time-consuming and energy-intensive, and easily deteriorate the quality of the final products (Humberto et al., 2001; Cohen and Yang, 1995). Scientists and technicists are looking for emerging food processing technologies to enable the production of safer, fresher and better quality foods with longer life for local and export markets. Among emergent new technologies ultrasound dehydration is very promising (Garcia-perez et al., 2007; Fuente-Blance et al., 2006).

Ultrasonic waves can produce a rapid series of alternative compressions and expansions, in a similar way to a sponge when it is squeezed and released repeatedly (Soria and Villamiel, 2010). The forces involved by this mechanism can be higher than surface tension which maintains the moisture inside the capillaries and microscopic channels of the fruit, and therefore accelerate moisture removal (Mulet et al., 2003). Ultrasonic waves can reduce viscosity (Greguss, 1963), minimize diffusion boundary layer thickness and remove moisture from solid liquid interfaces (Carcel et al., 2007a; Rodrigues and Fernades, 2007). Furthermore, changes in microstructure have been reported after ultrasonic treatments of fruits, which can additionally improve water removal (Fernandes et al., 2008a).

Ultrasonic waves are often used to strengthen osmotic dehydration (OD) of food (Carcel et al., 2007b; Riera et al., 2004). Simal et al. (1998) strengthened OD with ultrasound technology, and concluded that both dewater rate and moisture removal ratio increase significantly compared with traditional OD process. Lenart and Auslander (1980) reported that moisture diffusion rate would increase with the rise of ultrasonic power. Yet Floros and Liang (1994) indicated that the rise of moisture diffusion rate could slow down and eventually stop when ultrasonic power continued to increase to a large value. So the influence of ultrasound on increasing OD process is limited. The advantage of using ultrasound is that the process can be carried out at ambient temperature and no heating is required, reducing the probability of food degradation (Mason, 1996). But OD process can only remove part of water inside materials and totally dried products can't be obtained, even with the assistance of ultrasound. So in order to get dried products, further drying processes such as AD should be carried out after ultrasound assisted osmotic dehydration (UOD). The increase of water diffusivity and the reduction of overall drying time with UOD pretreatment have recently been reported about several fruits and vegetables (Fernandes and Rodrigues, 2008b; Duan et al., 2008; Garcia-Noguera et al., 2010).

This study has investigated the use of ultrasound as a helpful treatment to OD prior to AD of carrot. Hot-air dryings with UOD pretreatment (UOD-AD) were investigated. Weight reduction ratio (WRR) of UOD and the drying time and effective moisture diffusivity of the following AD were studied. Carotenoid content of carrot product and energy cost of UOD-AD process with different ultrasonic energy densities were also analyzed.

Materials and Methods

Materials    Fresh ripe carrots were obtained from a local special production base of National Agricultural Science Institute in the province of Henan, China and were stored in a refrigerator at 2 ∼ 4°C. Prior to the start of each experiment, carrots were cut into slices (30 ± 1 mm in diameter and 5 ± 0.1 mm in thickness) with a slicer and the size of each slice was measured using a vernier caliper. The moisture content of samples was determined in a vacuum oven at 70°C for 24 h (AOAC, 1990), and the initial moisture content of fresh carrot was 8.40 ∼ 8.45 g/g (dry base).

Major equipments    The main experimental equipments include: hot-air drying equipment (GZ-II, Tianli Co., China); digital ultrasonic generator (KQ5200DE, Shumei Co., China); water bath (HH-8,Huanyu Co., China); electronic balance (BT2235, precision 0.01 g, Hongheng Co., China); refractometer (WYT-4, Jingmi Co., China); high performance liquid chromatography system (Agilent 1260, Agilent Co., USA).

Experimental method    The osmotic solution with different osmosis concentrations was prepared by mixing food-grade sucrose with distilled water. Carrot slices of 50 g were immersed into glass beakers with osmotic solution and the ratio of samples to solution was 1:8 to avoid dilution effects. Then the beakers were placed inside the temperature-controlled water bath and the ultrasonic probe was put into the solution to carry out UOD process, and the influences of ultrasonic energy density (0, 0.11, 0.22, 0.33, 0.44, 0.55 and 0.66 W/mL) and sucrose solution concentration(20, 30, 40, 50, 60 and 70 °Brix) on WRR were studied, at the osmosis temperature of 50°C. Ultrasonic energy density (UED) was determined by dividing applied ultrasound power by volume of process object. Power ultrasound was applied at a fixed frequency of 22 kHz. When UOD pretreatment was completed, the dehydrated samples were taken out from the beakers. Excess solution on the surface of samples was removed with tissue paper rapidly and weighting was performed using the digital balance. AD processes were carried through after UOD pretreatment, with air temperatures of 40, 50, 60 and 70°C and air velocity of 0.5, 1 and 1.5 m/s, respectively. The samples were taken out and weighted at regular time intervals, and then put back for further drying immediately. The drying process was carried on until the sample weight remained constant. All experiments were performed in triplicates and average results were used for analysis.

WRR    is calculated as follows:

  

Where W₀ is the initial weight of samples (g) and W₁ is sample weight after UOD (g).

Moisture content (M_UO) after UOD is calculated based on the following equation:

  

Moisture content (M) in AD is calculated as follows (Sharma et al., 2005):

  

Where W is the changed mass of samples during AD process (g), and W_d is dry matter mass of samples (g).

Carrot slices were considered as infinite slab because the thickness of samples (0.005 m) was much less than the diameter (about 0.03 m). So the moisture ratio (MR) and D_eff were therefore calculated by the following equation (Crank, 1975):

  

Where D_eff is the effective moisture diffusion coefficient (m²/s), L is the half of slice thickness (m), M_e is the equilibrium moisture content, M₀ is the initial moisture content, M_t is moisture content at time of t and t is the time (s).

The equation above is evaluated numerically for Fourier number (F_o), and can be rewritten as (Sharma et al., 2005):

  

Therefore, the value of D_eff was calculated by the following equation:

  

The data were analyzed statistically using Origin 8.0 software.

Carotenoid content    The carotenoid extraction was determined with the method of Ma et al. (2008). The carotenoid analysis was carried out by HPLC method. An Alltech C18 column (250 mm × 4.6 mm, 5 µm, Deerfield, USA) was used at a column temperature of 25°C. The mobile phases were A: methanol and B: ethyl acetate (85:15, v/v) at a flow rate of 1 mL/min and detection at 450 nm. Injecting size was 10 µm. Methanol and ethyl acetate were of HPLC grade (Fisher Scientific, USA). Carotenoid standard was purchased from Sigma Co. (USA). All other reagents were of analysis grade. The carotenoid content was expressed as mg carotenoid per 100 g carrot on dry basis. All determinations were carried out in triplicates.

Energy cost    The energy costs of UOD and AD were measured using a power meter (DTST79, Dahua Co., China). The energy cost value of UOD-AD process was determined as the sum of energy cost for UOD and energy cost for the following AD.

Results and Discussion

Effect of UOD time on WRR    Carrot slices were treated by UOD with different times of 10, 20, 30, 40, 50 and 60 min respectively, with other operation parameters of ultrasonic energy density 0.44 W/mL, osmosis temperature 50°C and sucrose solution concentration 60 °Brix, respectively. The values of WRR are shown in Fig.1. As osmosis time increased, WRR rised rapidly, yet the increase of WRR slowed down after 30 min osmosis. The contents of dissoluble substance between two phases of osmosis system vary notably at the initial period of UOD, and the caused osmotic pressure difference is large enough to achieve high dewater rate. As osmosis going on, the osmotic pressure difference between carrot and solution decreases and results in obvious falling-down of water removal rate. So a UOD period of 30 min is chosen for further study of carrot dehydration. Fernandes and Rodrigues (2008b) and Nowacka et al.(2012) also carried out UOD no longer than 30 min, and obvious weight loss were observed.

Fig. 1.

WRR with different osmosis times at ultrasonic energy density 0.44 W/mL, osmosis temperature 50°C and solution concentration 60 °Brix

Effect of ultrasonic power on WRR    Adjusting ultrasonic energy densities of 0, 0.11, 0.22, 0.33, 0.44, 0.55 and 0.66 W/mL, UOD of carrot was carried out under the condition of osmosis time 30 min, osmosis temperature 50°C and solution concentration 60 °Brix, respectively. Fig.2 showed that the rise of ultrasonic energy density increases WRR within the same operating duration. The use of ultrasound during OD resulted in more weight loss of samples. Higher ultrasound power produces stronger cavitations' effects and strengthens mass transfer rate as a result (Rodrigues and Fernandes, 2007). Moreover, microcosmic tunnels produced by supersonic vibration could obviously accelerate moisture diffusive rate (Fernandes and Rodrigue, 2008b). Yet when ultrasonic energy density exceeded 0.44 W/mL, WRR began to increase tardily as ultrasonic energy density increased. Excessive ultrasound power in osmosis dehydration doesn't produce more microcosmic tunnels and sponge effects becomes steady (Carcel et al., 2007b), so mass transfer rate doesn't increase obviously when ultrasonic energy density is over 0.55 W/mL.

Fig. 2.

Effects of ultrasonic energy density on WRR under the conditions of osmosis time 30 min, osmosis temperature 50°C and solution concentration 60 °Brix

Effect of sucrose solution concentration on WRR    Carrot slices were put into sucrose solution with different concentrations of 20, 30, 40, 50, 60 and 70 °Brix successively, and UOD was carried out with other parameters of osmosis time 30 min, osmosis temperature 50°C and ultrasonic energy density 0.44 W/mL respectively. The results of WRR were shown in Fig.3. WRR increased with increase of sucrose content inside osmotic solution. Higher solution concentration would improve concentration gradient between solution and materials, and enhance diffusive rate and dewater ratio as a result (Teles et al., 2006). Yet osmotic solution with over-high concentration could bring about higher viscosity and lower diffusivity coefficient, causing rising-up of mass transfer resistance (Abraao et al., 2013). So sucrose concentration of osmotic solution in UOD process should not exceed 60 °Brix.

Analysis of variables (ANOVA) carried out to see the effects of process variables on the WRR values, which were shown in Table 1, revealed that ultrasonic energy density and osmotic concentration had significant effects on WRR at 1% level. The ANOVA result of WRR for a mathematical model in terms of operational parameters showed an excellent fit. The value of R² was 0.9784 and the resulting model was expressed as follows:

  

Fig. 3.

Effects of osmotic solution concentration on WRR with osmosis time 30 min, osmosis temperature 50°C and ultrasonic energy density 0.44 W/mL

Table 1. Analysis of variance for the effects of ultrasonic energy density and osmotic concentration of UOD on WRR

Factor	WRR
	Sum of squares	F value
Ultrasonic energy density, X1	0.1009*	15.699
Osmotic concentration, X2	0.0618*	24.041

*Significant at p < 0.01

Drying characteristics of UOD-AD    After UOD pretreatment with the condition of dehydration time 30 min and osmosis temperature 50°C, at different ultrasonic energy densities and solution concentrations, AD of carrot slices was carried out at the parameters of air temperature 60°C and air velocity 0.5 m/s. Drying curves were shown in Fig.4 and Fig.5, and total dehydration times were shown in Table 2. In Fig. 4, when the solution concentration was fixed as 60 °Brix, it's observed that with the increase of ultrasonic energy density, the subsequent AD time decreased obviously. The time for AD with ultrasonic energy density of 0.55 W/mL was nearly half of that with ultrasonic energy density of 0 W/mL, which means the application of ultrasound in OD pretreatment could shorten the following AD time obviously. Higher ultrasonic energy density in UOD reduces moisture content of carrot slices after osmosis as mentioned above, which reduces mass of water to be removed during AD process. Moreover, higher ultrasonic power could enlarge capillary size and increase microscopic tunnels (Fernandes and Rodrigue, 2008b), which enhance mass transfer rate during AD process.

Fig. 4.

Drying curves of AD with UOD pretreatment at different ultrasonic energy densities with the conditions of osmosis time 30 min, osmosis temperature 50°C, solution concentration 60 °Brix, drying temperature 60°C and air velocity 0.5 m/s

Fig. 5.

Drying curves of AD with UOD pretreatment at different solution concentrations with the conditions of osmosis time 30 min, osmosis temperature 50°C, ultrasonic energy density 0.44 W/mL, drying temperature 60°C and air velocity 0.5 m/s

Table 2. the influences of ultrasonic energy density and osmotic concentration of UOD pretreatment on drying time and effective moisture diffusivity of the following hot-air drying (osmosis time 30 min, osmosis temperature 50°C, drying temperature 60°C and air velocity 0.5 m/s)

Ultrasonic energy density (W/mL)	Osmotic concentration (°Brix)	AD time (min)	Total processing time (min)	Deff × 109 (m2/s)
0	20	423	453	1.825
0	40	392	422	1.965
0	60	374	404	2.180
0.11	20	335	365	2.519
0.11	40	313	343	2.725
0.11	60	304	334	2.905
0.22	20	311	341	3.080
0.22	40	277	307	3.210
0.22	60	255	285	3.378
0.33	20	288	318	3.460
0.33	40	247	277	3.610
0.33	60	223	253	3.800
0.44	20	272	302	3.980
0.44	40	225	255	4.130
0.44	60	193	223	4.306
0.55	20	258	288	4.110
0.55	40	204	234	4.320
0.55	60	187	217	4.526

(osmosis time 30 min, osmosis temperature 50°C, drying temperature 60°C and air velocity 0.5 m/s)

Figure 5 showed that AD time decreases with the increase of osmosis concentration in UOD pretreatment with ultrasonic energy density of 0.44 W/mL. The time of AD with the osmosis concentration of 60 °Brix was around 190 min, which is much shorter than the drying time of 270 min with the osmosis concentration of 20 °Brix. Higher osmosis concentration removes more moisture from samples, and hence lowers dewater quantity and shortens AD time.

The AD times, the total processing times and the values of D_eff of UOD-AD were shown in Table 2. Total processing time equals the sum of UOD time and the following AD time. The drying time of AD without UOD pretreatment was 340 min at the conditions of drying temperature 60°C and air velocity 0.5 m/s. The total dehydration time of AD with OD pretreatment at the ultrasonic energy density of 0 W/mL was obviously longer than that of direct AD, which means OD pretreatment without ultrasound application prolongs dehydration time instead of shortening it. Without ultrasound assistance, OD can only remove small part of water inside samples and WRR is low, so more moisture need to be removed in the following AD process. In addition, osmotic solution could produce a thin sucrose membrane on the surface of samples, which causes the increase of mass transfer resistance (Garcia-noguera et al., 2010). So AD with OD pretreatment alone is time-comsuming and unadvisible. When ultrasound technology is applied, AD time decreased significantly. Ultrasound can produce strong turbulence on the surface of samples which could prevent formation of the membrane, and is beneficial for water removal to air flow, and reduces the total processing time as a result.

The values of D_eff increased obviously with the increase of ultrasonic energy density and osmotic concentration. For example, at the osmotic concentration of 60 °Brix, the values of D_eff increased from 2.18 × 10⁻⁹ m²/s to 4.52 × 10⁻⁹ m²/s when ultrasonic energy density increased from 0 W/mL to 0.55 W/mL. Increased D_eff value has been associated with formation of microscopic channels in the intercellular tissue of fruits (Fernandes et al., 2008a; Carcel et al., 2007). Such formation may be due to the separation or disruption of cells caused by the combined effects of cavitation and osmotic pressure (Garcia-Noguera et al., 2010). This positive effect makes it easier for water to diffuse during AD process. As a result, drying time is shortened and D_eff value is increased.

According to the results of ANOVA shown in Table 3, it's revealed that ultrasonic energy density and osmotic concentration had significant effects on the drying time and D_eff value. The mathematical models for drying time t₁ and D_eff,1 of UOD-AD at the conditions of drying temperature 60°C and air velocity 0.5 m/s are shown as follows.

  

  

Table 3. Analysis of variance for the overall effects of ultrasonic energy density and osmotic concentration of UOD pretreatment on drying time and effective moisture diffusivity of hot air drying

Factor	Drying time, t1		Deff,1
	Sum of squares	F value	Sum of squares	F value
Ultrasonic energy density, X1	66592.94*	164.044	11.703*	4218.927
Osmotic concentration, X2	10584.78*	65.186	0.3755*	338.421

*Significant at p<0.01

Figures 6 and 7 showed the drying curves of AD at different air temperatures and air velocities after UOD pretreatment with the parameters of osmosis time 30min, ultrasonic energy density 0.44 W/mL, solution concentration 60 °Brix and osmosis temperature 50°C, respectively. The initial moisture content of samples for AD was 4.47 ∼ 4.53 g/g (dry base) since UOD pretreatment of the same operation parameters was carried out before AD. It's shown that the drying time are 427, 234, 193 and 145 min when the drying temperatures are 40, 50, 60 and 70°C, respectively. With the increase of air temperature, heat transfer rate rises markedly and enhances mass transfer rate and water evaporation rate consequently. Yet air velocity has little effects on the AD characteristics of pretreated samples. The influences of drying temperature and air velocity on the following AD are similar to that of the two parameters on direct AD of fruits and vegetables reported in other literatures (Srikiatden and Roberts, 2006; Orikasa et al., 2006; Maskan, 2001).

Fig. 6.

Drying curves of UOD-AD at different drying temperatures with the conditions of osmosis time 30 min, ultrasonic energy density 0.44 W/mL, solution concentration 60 °Brix , osmosis temperature 50°C and air velocity 0.5 m/s

Fig. 7.

Drying curves of UOD-AD at different air velocities with the conditions of osmosis time 30 min, ultrasonic energy density 0.44 W/mL, solution concentration 60 °Brix, osmosis temperature 50°C, and drying temperature 60°C

The drying curves of AD without pretreatment are shown in Fig. 8. The total times of AD and UOD-AD are shown in Table 4. Based on the figures and the table, it's concluded that UOD pretreatment prior to AD could reduce total processing time approximately 80 ∼ 130 min compared with direct AD. Lower initial moisture content and the change of microstructure caused by ultrasonic osmosis before AD are the two main factors to shorten total processing time. So UOD is helpful to increase total dehydration rate and save drying time. The D_eff values of AD and UOD-AD were shown in Table 4. At all levels of drying temperatures and air velocities, the values of D_eff of UOD-AD are higher than that of AD at the same drying conditions So it's concluded that UOD pretreatment has positive effect on enhancing water diffusivity and shortening drying time at the conditions of drying temperatures in 40 - 70°C and air velocities in 0.5 - 1.5 m/s.

Fig. 8.

Drying curves of AD without UOD pretreatment at different drying temperatures at air velocity of 0.5 m/s

Table 4. Drying time and effective moisture diffusivity of hot-air dryings with UOD pretreatment and without pretreatment

Drying temperature (°C)	Air velocity (m/s)	UOD-AD			Direct AD
		AD time (min)	Total processing time (min)	Deff × 109 (m2/s)	AD time (min)	Deff × 109 (m2/s)
40	0.5	427	457	1.807	568	1.334
40	1	415	445	1.952	545	1.490
40	1.5	398	428	2.081	527	1.610
50	0.5	234	264	3.411	388	2.094
50	1	221	251	3.574	370	2.210
50	1.5	205	235	3.713	355	2.320
60	0.5	193	223	4.306	340	2.432
60	1	180	210	4.536	325	2.640
60	1.5	168	198	4.722	308	2.870
70	0.5	145	175	5.455	255	3.344
70	1	132	162	5.682	240	3.638
70	1.5	119	149	5.817	226	3.898

(Process parameters of UOD pretreatment are osmosis time 30 min, ultrasonic energy density 0.44 W/mL, solution concentration 60 °Brix and osmosis temperature 50°C)

ANOVA was carried out and the results of the overall effects of drying temperature and air velocity on drying time and effective moisture diffusivity of UOD-AD were shown in Table 5. It's revealed that drying temperature and air velocity had significant effects on the drying time and D_eff value. The mathematical models for drying time t₂ and D_eff,2 of AD with UOD pretreatment are shown as follows.

  

Table 5. Analysis of variance for the overall effects of drying temperature and air velocity on drying time and effective moisture diffusivity of UOD-AD process

Factor	Drying time, t2		Deff, 2
	Sum of squares	F value	Sum of squares	F value
Drying temperature, X3	136851.6*	26921.62	22.1354*	6700.765
Air velocity, X4	1487.167*	438.836	0.2305*	104.644

*Significant at p<0.01

Carotenoid content of carrot product in UOD-AD    Carotenoid contents of carrot product after UOD-AD at different ultrasonic energy densities were shown in Fig.9. The UOD pretreatment parameters were osmosis time 30 min, solution concentration 60 °Brix and osmosis temperature 50°C, respectively. AD was carried out at drying temperature 60. and air velocity 0.5 m/s, respectively. Carotenoid content of carrot product improved when ultrasonic energy density increased from 0 to 0.22 W/mL. UOD pretreatment could lead to water removal from samples to osmosis solution and a portion of carotenoid compounds diffuse into the solution at the same time, resulting in a loss of carotenoid content during UOD pretreatment. On the other hand, carotenoid is thermal-sensitive and oxidizable, and easily degraded and isomerized during drying process (Klieber and Bagnato, 1999). So both UOD pretreatment and following AD process give rise to the losing of carotenoid compounds. When ultrasonic energy density was low in UOD pretreatment, mass transfer was weak and quantity of carotenoid moved from samples to osmosis solution was small. Yet the drying time of the following AD was too long, and resulted in serious carotenoid degradation. As ultrasound power increased, the drying time of following AD was shortened obviously, causing less carotenoid isomerization and oxidation and reducing carotenoid loss as a result. Yet when ultrasonic energy density increased from 0.22 to 0.55 W/mL, carotenoid content of carrot product slightly reduced in spite of the shortening of following AD time. Higher ultrasonic energy density produces stronger cavitations' effects and turbulence as mentioned above, and arouses physical damage of organization structure and increase of cell membrane permeability (Gabaldon-Leyva et al., 2007). Such effects and transformation impel more carotenoid compounds moved from carrot organization into osmosis solution consequently. So this negative influence of UOD pretreatment on carotenoid content counteracts the positive influence of short drying time of UOD-AD on carotenoid content, and causes slight decline of carotenoid content in dried product with the increase of ultrasonic energy density in UOD pretreatment. The highest carotenoid content value came from UOD-AD at ultrasonic energy density of 0.22 W/mL, and this result indicated that ultrasonic energy density in UOD pretreatment should not be very low or excessive in order to protect carotenoid compounds in carrot product of UOD-AD.

Fig. 9.

Carotenoid content of carrot product of UOD-AD at different ultrasonic energy densities with the conditions of osmosis time 30 min, solution concentration 60 °Brix, osmosis temperature 50°C, drying temperature 60°C and air velocity 0.5 m/s

Compared with carotenoid contents of carrot products of direct AD and UOD-AD, it's illustrated in Fig. 9 that the carotenoid content values of UOD-AD were higher than that of direct AD. The highest carotenoid content value of UOD-AD was 8.7% higher than the value of AD. So it could be concluded that UOD pretreatment is beneficial to improving carotenoid content of carrot products in AD and reducing drying time as well. The carotenoid content in OD-AD without ultrasound assistance was lower than that in direct AD. The case may be caused by too long drying period of OD-AD without ultrasound assistance.

Energy cost of UOD-AD    Energy cost values of UOD-AD at different ultrasonic energy densities were listed in Table 6. The UOD pretreatment parameters were osmosis time 30 min, solution concentration 60 °Brix and osmosis temperature 50°C, respectively. AD was carried out at drying temperature 60°C and air velocity 0.5 m/s. Since UOD pretreatment with higher ultrasonic energy density results in less drying time of the following AD, energy cost decreases with the increase of ultrasonic energy density in UOD-AD. The energy cost values ranged from 2.05 to 1.38 kW·h for UOD-AD at ultrasonic energy densities of 0.11 - 0.55 W/mL, which were lower than the value of 2.17 kW·h for direct AD. The smallest value was 36% less than energy cost for direct AD. So it is confirmed that UOD pretreatment can reduce energy cost of AD process significantly. Because of longer drying time, the energy cost of OD-AD without ultrasound assistance was 14.7% higher than that of direct AD, which means OD pretreatment has to increase energy cost of AD unexpectedly.

Table 6. Energy cost analysis of UOD-AD at different ultrasonic energy densities

Ultrasound power (W)	Energy cost, (kW·h)
0	2.49
50	2.05
100	1.75
150	1.56
200	1.39
250	1.38
AD	2.17

(osmosis time 30 min, solution concentration 60 °Brix and osmosis temperature 50°C for UOD pretreatment and drying temperature 60°C and air velocity 0.5 m/s for AD process)

Conclusions

Compared with traditional OD process, ultrasound could increase mass transfer rate inside and outside carrot slices during OD and reduce dehydration time obviously. Yet too high ultrasonic energy density and solution concentration could not enhance moisture diffusivity significantly. Through the experiments of UOD-AD of carrot, it's concluded that the UOD pretreatment parameters including ultrasonic energy density and solution concentration and AD parameters including drying temperature and air velocity have significant effects on total processing time and effective moisture diffusivity. OD pretreatment without ultrasound assistance could prolong total processing time, yet the application of ultrasound could increase dewater efficiency and the effective moisture diffusivity at AD stage and reduce processing time obviously. And UOD pretreatment could improve carotenoid content of carrot product and reduce energy cost of the dehydration process compared with direct AD. Therefore UOD pretreatment is an interesting and effective methodology complementary to traditional AD technology.

Acknowledgement    This work was financially supported by the National Natural Science Foundation of China (31171723 and 11004049), and the Science and Technology Research Program of Henan Province (12A210005).

References

•  Abraao,  A.S.,  Lemos,  A.M.,  Vilela,  A.,  Sousa,  J.M., and  Nunes,  F.M. (2013). Influence of osmotic dehydration process parameters on the quality of candied pumpkins. Food Bioprod. Process., In press.
• AOAC. (1990). Official method of analysis. Arlington, VA: Association of Official Analytical Chemists. No.934.06.
•  Carcel,  J.A.,  Garcia,  P.J.V.,  Riera,  E., and  Mulet,  A. (2007a). Influence of high-intensity ultrasound on drying kinetics of persimmon. Dry. Technol., 25, 85-193.
•  Carcel,  J.A.,  Benedito,  J.,  Rossello,  C., and  Mulet,  A., (2007b). Influence of ultrasound intensity on mass transfer in apple immersed in a sucrose solution. J. Food Eng., 78, 472-479.
•  Cohen,  J.S. and  Yang,  T.C.S. (1995). Progress in food dehydration. Trends Food Sci. Tech., 6, 20-25.
•  Crank,  J. (1975). Mathematic of diffusion. 2nd ed. Oxford University Press, London.
•  Duan,  X.,  Zhang  M.,  Li  X., and  Mujumdar,  A.S. (2008). Ultrasonically enhanced osmotic pretreatment of sea cucumber prior to microwave freeze drying. Dry. Technol., 26, 420-426.
•  Fernandes,  F.A.N.,  Gallao,  M.I., and  Rodrigues,  S. (2008a). Effect of osmotic dehydration and ultrasound pre-treatment on cell structure: Melon dehydration. LWT-Food Sci. Technol., 41, 604-610.
•  Fernandes,  F.A.N. and  Rodrigues,  S. (2008b). Application of ultrasound and ultrasound-assisted osmotic dehydration in drying of fruits. Dry. Technol., 26, 1509-1516.
•  Floros,  J.D. and  Liang,  H. (1994). Acoustically assisted diffusion through membranes and Biomaterials. Food Technol., 10, 79-84.
•  Fuente-Blance,  S.D.,  Sarabia,  E.R.F.,  Acosta-Aparicio,  V.M.,  Blanco,  B.A., and  Gallego-Juarez,  J.A. (2006). Food drying process by power ultrasound. Ultrasonics, 44, 523-527.
•  Gabaldon-Leyva,  C.A.,  Quintero-Ramos,  A.,  Barnard,  J.,  Balandran-Quintana,  R.R.,  Talamas-Abbud,  R., and  Jimenez-Castro,  J. (2007). Effect of ultrasound on the mass transfer and physical changes in brine bell pepper at different temperatures. J. Food Eng., 81, 347-379.
•  Garcia-Noguera,  J.,  Oliveira,  F.I.P.,  Gallao,  M.I.,  Weller,  C.L.,  Rodrigues,  S., and  Fernandes,  F.A.N. (2010). Ultrasound-assisted osmotic dehydration of strawberries: effect of pretreatment time and ultrasonic frequency. Dry. Technol., 28, 294-303.
•  Garcia-perez,  J.V.,  Crecel,  J.A.,  Benedito,  J., and  Mulet,  A. (2007). Power ultrasound mass transfer enhancement in food drying. Food Bioprod. Process., 85, 247-254.
•  Greguss,  P. (1963). The mechanism and possible applications of drying by ultrasonic irradiation. Ultrasonics, 2, 83-86.
•  Humberto,  V.M.,  Marcela,  G.N., and  Gustavo,  V.B.C. (2001). Advances in dehydration of foods. J. Food Eng., 49, 271-289.
•  Klieber,  A. and  Bagnato,  A. (1999). Colour stability of paprika and chilli powder. Food Aust, 51, 592-596.
•  Lenart,  I. and  Auslander,  D. (1980). The Effect of Ultrasound on Diffusion Through Membranes. Ultrasonics, 9, 6-218.
•  Ma,  W.P.,  Ni,  Z.J.,  Li,  H., and  Chen,  M. (2008). Changes of the main carotenoid pigment contents during the drying processes of the different harvest stage fruits of Lycium barbarum L.. Agr Sci China, 7, 363-369.
•  Maskan,  M. (2001). Drying , Shrinkage and rehydration characteristics of kiwifruits during hot air and microwave drying. J. Food Eng., 48, 177-182.
•  Mason,  T.J.,  Paniwnyk,  L., and  Lorimer,  J.P. (1996). The use of ultrasound in food technology. Ultrason. Sonochem., 3, 253-256.
•  Mulet,  A.,  Carcel,  J.A.,  Sanjuan,  N., and  Bon,  J. (2003). New food drying technologies-use of ultrasound. Food Sci. Technol. Int., 9, 215-221.
•  Nowacka,  M.,  Wiktor,  A.,  Sledz,  M.,  Jurek,  N., and  Witrowa-Rajchert,  D. (2012). Drying of ultrasound pretreated apple and its selected physical properties. J. Food Eng., 113, 427-433.
•  Orikasa,  T.,  Wu,  L.,  Shiina,  T., and  Tagawa,  A. (2006). Drying characteristics of kiwifruit during hot air drying. J. Food Eng., 85, 303-308.
•  Riera,  E.,  Golas,  Y.,  Blanco,  A.,  Blasco,  M., and  Mulet,  A. (2004). Mass transfer enhancement in supercritical fluids extraction by means of power ultrasound. Ultrason. Sonochem., 11, 241-244.
•  Rodrigues,  S. and  Fernades,  F.A.N. (2007). Use of ultrasound as pretreatment for dehydration of melons. Dry. Technol., 25, 1791-1796.
•  Sharma,  G.P.,  Verma,  R.C., and  Pathare,  P.B (2005). Thin-layer infrared radiation drying of onion slices. J. Food Eng., 67, 361-366.
•  Simal,  S.,  Bendito,  J.,  Sanchez,  E.S., and  Rossello,  C. (1998). Use of ultrasound to increase mass transport rates during osmotic dehydration. J. Food Eng., 36, 323-336.
•  Soria,  A.C. and  Villamiel,  M. (2010). Effect of ultrasound on the technological properties and bioactivity of food. Trends Food Sci. Tech., 21, 323-331.
•  Srikiatden,  J. and  Roberts,  J.S. (2006). Measuring moisture diffusivity of potato and carrot, (core and cortex)during convective hot air and isothermal drying. J. Food Eng., 74, 143-152.
•  Teles,  U.M.,  Fernandes,  F.A.N.,  Rodrigues,  S.,  Lima,  A.S.,  Maia,  G.A., and  Figueiredo,  R.W. (2006). Optimization of osmotic dehydration of melons followed by air-drying. Int. J. Food Sci. Tech., 41, 674-680.