Release behavior of allyl sulfide from cyclodextrin ...

Bioscience, Biotechnology, and Biochemistry

ISSN: 0916-8451 (Print) 1347-6947 (Online) Journal homepage: http://www.tandfonline.com/loi/tbbb20

Release behavior of allyl sulfide from cyclodextrin inclusion complex of allyl sulfide under different storage conditions Thi Van Anh Nguyen & Hidefumi Yoshii To cite this article: Thi Van Anh Nguyen & Hidefumi Yoshii (2018) Release behavior of allyl sulfide from cyclodextrin inclusion complex of allyl sulfide under different storage conditions, Bioscience, Biotechnology, and Biochemistry, 82:5, 848-855, DOI: 10.1080/09168451.2018.1440173 To link to this article: https://doi.org/10.1080/09168451.2018.1440173

Published online: 02 Mar 2018.

Submit your article to this journal

Article views: 80

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=tbbb20

Bioscience, Biotechnology, and Biochemistry, 2018 VOL. 82, NO. 5, 848–855 https://doi.org/10.1080/09168451.2018.1440173

Release behavior of allyl sulfide from cyclodextrin inclusion complex of allyl sulfide under different storage conditions Thi Van Anh Nguyena,b,c and Hidefumi Yoshiia,b a

Department of Applied Biological Science, Kagawa University, Kagawa, Japan; bApplied Bioresource Sciences Department, The United Graduate School of Agricultural Sciences, Ehime University, Ehime, Japan; cFaculty of Engineering and Food Technology, Hue University of Agriculture and Forestry, Hue University, Hue, Vietnam

ABSTRACT

The stability of allyl sulfide, an organosulfur compound present in garlic oil, in its α-, β-, and γ-cyclodextrin inclusion complexes was investigated under various storage conditions. The complexes of cyclodextrins and allyl sulfide were prepared by spray drying. The storage temperature, relative humidity, and initial moisture content of the inclusion complex had different effects on the release rate of allyl sulfide. Allyl sulfide in α-cyclodextrin complexes had a lower release rate than in β- and γ-cyclodextrin complexes at 100 °C and at 50 °C under 6, 40, 54, and 73% relative humidity. The initial moisture content affected only the release rate of allyl sulfide from α-cyclodextrin complexes. The release behavior of allyl sulfide can be correlated with the first-order release rate equation with a normal Gaussian distribution of free energy of activation of release rate constant. The results indicated α-cyclodextrin is a suitable material for controlled release of allyl sulfide.

Microencapsulation of flavors in cyclodextrins (CDs) is an efficient technique that protects them from environmental factors such as light, heat, and oxygen, and also enhances their chemical and physical properties. Hydrophobic flavors could be encapsulated in the cavity of CDs if the molecules have an appropriate size, geometry, and polarity match with the cavity. Owing to the differences in cavity diameter (0.57, 0.78, and 0.95 nm) and solubility (14.5, 1.85, and 23.2 g/100 g H2O), α-, β-, and γ-CDs exhibit the different encapsulation abilities [1]. The inclusion complexes of CDs and flavors can be prepared by a precipitation [2], kneading [3], and spray drying [4]. Among these encapsulation methods, spray drying is usually used because it is efficient, inexpensive, and leads to the formation of dry flavor powder [5]. Flavor is an important sensory key for food quality. Allyl sulfide, a highly volatile sulfur-containing component of garlic oil, is responsible for the characteristic garlic flavor. Commercially, garlic essential oil is used in its diluted form with vegetable oil as the diluting solvent [6]. Medium chain triglycerides (MCT) oil (from either coconut oil or palm seed) is commonly used as the commercial water-insoluble solvent [7]. During the encapsulation process, either MCT or olive oils may be used as solvents for diluting alpha-tocopherol [8,9]. Control of flavor release is essential in the food and flavoring industry because it determines the shelf-life

CONTACT Hidefumi Yoshii

[email protected]

© 2018 Japan Society for Bioscience, Biotechnology, and Agrochemistry

ARTICLE HISTORY

Received 4 October 2017 Accepted 6 February 2018 KEYWORDS

Stability; molecular encapsulation; cyclodextrin; spray drying; allyl sulfide

and consumer acceptability of the flavored product and controlled flavor applications. Madene et al. [10] have summarized the flavor release mechanism caused by diffusion, degradation, swelling, and melting. The release of flavor can be investigated by modeling the release during eating, under different storage conditions (at different temperatures, relative humidity (RH), and pH values), or in a real food matrix. Kant et al. [11] have demonstrated that β-CD can be used to modify flavor delivery in both model and real systems. Shiga et al. [12] evaluated the release characteristics of allyl isothiocyanate (AITC) encapsulated in α-, β-, and γ-CD and found that α-CD was good material for controlling release of AITC. The release of AITC from the inclusion complex powders was a result of the replacement of AITC with water molecules. The stability of the key volatile odor of durian spray-dried powder during accelerated storage was evaluated by Chin et al. [13]. These authors found that the release rate was dependent on RH. Shiga et al. [4] investigated the release of d-limonene and ethyl n-hexanoate from spray-dried powders of the blended encapsulant of CD and gum arabic at 50 °C and 75% RH. They determined that the release of flavor from spraydried powders depends on the nature of flavor molecule and the composition of the encapsulant. Capozzi et al. [14] prepared an ethylene/α-CD inclusion complex using a saturated solution method and regulated its release rate at 30 and 50 °C, and 75 and 97% RH. Thus, the study

BIOSCIENCE, BIOTECHNOLOGY, AND BIOCHEMISTRY

showed that the release rate of volatile flavor from spraydried powder was markedly affected by the RH. Shiga et al. [15] and Wang et al. [16] reported that the release behavior of flavor from cyclodextrin inclusion complex powders could be correlated with the Avrami equation. However, in some flavors, the release rate decreased very quickly at the beginning and then stabilized. The stability of ethyl eicosapentaenoate from CDs was correlated with a model of statistical distribution activation energy by Yoshii et al. [17], in which they used the first-order kinetic equation with a normal Gaussian distribution of free activation energy or Gibbs energy of activation. This original equation was proposed by Kawamura et al. [18] when they estimated the thermal inactivation of immobilized α-chymotrypsin. This equation was also applied for estimating the oxidation of spray-dried linoleic acid powder by Ishido et al. [19]. Wang et al. [20] investigated the in vitro release profile of garlic oil (GO) from a GO/β-CD complex in an acidic dissolution medium (pH 1.5) at 37 °C and reported an initial fast release phase followed by a delayed release period. Furthermore, Ho et al. investigated the release properties of a CO2-α-cyclodextrin complex powder prepared by solid encapsulation (water activity, aw ≈ 0.95) followed by moisture removal using silica gel and CaCl2 desiccants during post-dehydration. The release depended on RH, liquid environment (water or oil), and packing method (vacuum or normal) [21]. Our research focused on the release behavior of allyl sulfide from its α-, β-, and γ-CD inclusion complex powders at different RH, storage temperatures, and initial moisture content (IMC) of the inclusion complex powders. Their release behaviors were correlated using a release rate constant with a normal Gaussian distribution of the free energy.

Materials and methods α-, β-, and γ-CD were purchased from Wacker Chemical Corp (USA) and allyl sulfide (98% purity, density 0.89 g/ mL) was obtained from Tokyo Chemical Industry Co., Ltd. MCT oil and anisole (methoxybenzene) were bought from Riken Vitamin Co., Ltd. and Wako Pure Chemical Industries, Ltd., respectively. All other chemicals were of analytical grade and obtained from Wako Pure Chemical Industries, Ltd. (Tokyo, Japan). Table 1. Initial moisture content of the cyclodextrin inclusion complexes. Initial moisture content (%) Effect of temperature (Figures 1–3) 7.8 ± 2.04 10.7 ± 0.14 8.8 ± 0.51

Preparation of the inclusion complex powders Allyl sulfide diluted in MCT oil (concentration 10000 ppm; 3 wt%), CD (α, β, or γ-CD; 27 wt%), and water (70 wt%) were mixed using a Polytron homogenizer (PT-6100; Kinematica, Littau, Switzerland) at 1 × 104 rpm for 3 min. The mixture was incubated at 30 °C for 8 h at a shaking speed of 150 rpm using a shaker (NTS-1300; Eyela). The mixtures were spray dried in a B-290 Mini Spray Dryer (Büchi). The spray drying conditions were as follows: the atomizer-type nozzle had a feed flow rate of 10 mL/min, air flow rate of 35 m3/h, inlet air temperature was maintained at 140 °C and the outlet air temperature was 61–65 °C [22]. The inclusion complex powders (spray-dried powders) were stored at −30 °C in nitrogen-filled bags until required for the experiments. The IMC of the inclusion complexes of allyl sulfide (diluted in MCT oil) and α-, β-, and γ-CD is shown in Table 1. To obtain inclusion complexes with low moisture content, the powders were vacuum dried at 80 °C for 24 h under a pressure of 40 Pa. The moisture content of the α-, β-, and γ-CD inclusion complexes after vacuum drying treatment was 2.7, 1.8, and 2.7 wt%, respectively. The retention of allyl sulfide in the inclusion complexes treated by vacuum drying was also analyzed. It was found that there was no difference in allyl sulfide retention before and after the vacuum drying treatment. Initial moisture content of the inclusion complex powders The IMC of the inclusion complex powder was determined by using a Mettler Toledo HB43-S Compact Halogen Moisture Analyzer. Samples (1 ± 0.005 g) were put in an aluminum plate and heated at 160 °C. The IMC was presented as an average of triplicate measurements. Particle size of inclusion complex powders

Materials

Type of inclusion complex α-CD β-CD γ-CD

849

Effect of RH (Figure 4) 8.9 ± 1.05 11.3 ± 0.91 8.8 ± 0.07

The size distribution of particles in complex powders was measured by using a laser scattering particle size analyzer (SALD-7100; Shimadzu Corporation, Kyoto, Japan) installed with a batch sample cell. The inclusion complex powder was dispersed into 2-methyl-1-propanol and this dispersed solution was pipetted into a cell containing 2-methyl-1-propanol for measuring the size distribution of particle in the complex powders. Each sample was analyzed in triplicate. The volume-based diameter (D43) was regarded as the mean diameter. Stability of allyl sulfide in the inclusion complex powders To investigate the release of allyl sulfide from the CD inclusion complexes, about 0.1 g each of α-, β-, and γ-CD spray-dried powders was weighed in a Mighty vial (ø24 × 50 mm) and kept at 100 °C. In order to determine

850

T. V. A. NGUYEN AND H. YOSHII

the stability of allyl sulfide in MCT oil, 0.01 g of 10000 ppm allyl sulfide in MCT oil was used in the same vessel. The effect of storage temperature on the stability of allyl sulfide in the α-CD complex was investigated at 60, 80, and 100 °C. The experiments were performed as described above. The effect of RH and IMC on the release timecourse of allyl sulfide from the CD inclusion complex powders during the storage period was investigated at 50 °C. About 0.1 g of the spray-dried powders was weighed in a Mighty vial (ø24 × 50 mm) and placed in a plastic box containing silica gel, a saturated salt solution of potassium carbonate, sodium bromide, or sodium chloride to maintain 6, 40, 54, and 73% RH, respectively. These boxes were covered with a punched aluminum foil and kept at 50 °C in an oven [23,24]. Two types of α-, β-, and γ-CD inclusion complex powders with low IMC of 2.7, 1.8, and 2.7 wt%, and high IMC of 8.9, 11.3, and 8.8 wt%, respectively, were used for the experiments.

where C is the flavor concentration in the inclusion complex powder after storage (mg/g), C0 is the initial flavor concentration (mg/g), k10 is the release rate constant at an average ∆G (1/day), t is time (days), k1 is the release rate constant (1/day), R is ideal gas constant (J/(mol.K)), T is the absolute temperature (K), and σ is the standard deviation of the ∆G distribution of the release rate constant (−). Flavor retention C/C0 can be calculated as a function of time t at given values of the standard deviation σ and average release rate constant k10. k10 was assumed to be the first-order release rate constant at time 0 (day). The initial value of k10 was determined from the slope of a tangent to the curve of the experimental values at t = 0 (day). Then, by using program solver in Microsoft Excel, the release rate constant k10 and σ were determined so as to fit with the experimental data.

Retention of allyl sulfide in the inclusion complex powders

Figure 1 shows the release behavior of allyl sulfide from the α-, β-, and γ-CD inclusion complex powders and MCT oil (10000 ppm, 0.01 g oil) during five weeks’ storage at 100 °C. It was found that the release rate of allyl sulfide from MCT oil was markedly faster than that from the CD inclusion complex powders. This result indicated that the CD inclusion complex powders were ideal for the controlled release of allyl sulfide. Remarkably, in the first two weeks, nearly 90% of allyl sulfide in the β- and γ-CD powders was released, while 50% of the allyl sulfide still remained in the α-CD inclusion complex spray-dried powder. The solid lines shown in Figure 1 were calculated using Equation (1). As shown in Figure 1, the release time-course of allyl sulfide from the α-, β-, and γ-CD

Correlation of allyl sulfide release The flavor release from the spray-dried powders during storage was considered as a type of relaxation phenomenon in amorphous glass compounds containing numerous oil droplets of different sizes. In order to understand the flavors the release time course, the following equation was used to simulate flavor release behavior of the inclusion complex powders using a Gaussian distribution of the free energy of activation (∆G) or Gibbs energy of activation: [18] ∞

𝜙=

�

C RT =√ exp − R2 T lnk1 − lnk10 ∫ C0 2𝜋𝜎 −∞ (1) � � � � � � 2 + 𝜎 2 ∕2R2 T 2 ∕2𝜎 2 exp −k1 t d lnk1 � 2

Stability of allyl sulfide in the inclusion complex powders

Allyl sulfide retention ( )

The method for extracting and analyzing allyl sulfide from the CD inclusion complex was as reported by Nguyen and Yoshii [22]. In brief, samples were mixed with 2 mL of distilled water and 3 mL of hexane (containing anisole as an internal standard) for 3 min. The mixture was then heated to 90 °C for 30 min (interval 10 min, mixed 3 min). After centrifugation (3,000 rpm for 10 min), 1.0 μL of the supernatant was injected into the Gas Chromatography with Flame Ionization Detector (GC-FID) and the amount of total flavor was measured. All samples were analyzed twice and the data were presented as an average. The retention of allyl sulfide (ϕ) was defined as the ratio of the mean value of allyl sulfide concentration in the spray-dried powders after storage to the mean value of the allyl sulfide concentration in the initial spray-dried powders.

Results and discussion

1.0

0.5

0.0

0

7

14

21

28

35

Storage time (days)

Figure 1. Release behavior of allyl sulfide from its α-CD (), β-CD (■), and γ-CD (▲) inclusion complex powders and MCT oil (10000 ppm) (×) at 100 °C. The solid lines were calculated by Equation (1).


The release of allyl sulfide from its α-CD inclusion complex powder at different temperatures (60, 80, and 100 °C) is shown in Figure 2. It was found that allyl sulfide remained encapsulated in the α-CD at 60 and 80 °C for four weeks. At 100 °C, the shelf-life of the inclusion complex was more than two weeks. By using Equation (1), allyl sulfide release behavior was plotted as solid lines. From equation (1), k10 was calculated at different temperatures and the Arrhenius plot of ln k10 vs. T−1 was obtained (Figure 3). The activation energy for allyl sulfide release from α-CD inclusion complex powders was approximately 183 kJ/mol. Yoshii et al. [25] have reported that the release rate constant of (–)-menthol from β- to γ-CD complexes increase with an increase in temperature. Reineccius et al. [27] have also observed a strong temperature dependence of the flavor/CD binding. These authors also discussed the

Table 2. Release rate constants, k10, and standard deviations, σ, of the distribution of ∆G at 100 °C. Type α-CD β-CD γ-CD

k10 (day−1) 5.95 × 10−2 7.51 × 10−1 7.51 × 10−1

σ 0.8 0.99 0.99

1.0

Allyl sulfide retention ( )

Effect of temperature on the release behavior of allyl sulfide from the α-CD inclusion complex powders

thermal and oxidative stability of the α- and β-CD complexes of major garlic constituents. They showed that diallyl disulfide and diallyl trisulfide were suitably encapsulated at a higher concentration because of their higher log P value, and just a small decrease in the relative concentration was observed for all sulfide compounds in the presence of air or argon at 90 °C [28]. Soottitantawat et al. [29] showed that the activation energies of l-menthol release from the emulsified l-menthol spray-dried powders at 75 and 83% RH were 140 and 48 kJ/mol, respectively. Zhu et al. [30] have reported that the activation energy of methyl acetate from its β-CD inclusion complex powder was 259 kJ/mol. These results indicate that the activation energy of flavor release from its CD inclusion complexes is usually higher than that from spray-dried powders.

0.5

0.0

0

7

14

21

28

35

Storage time (days)

Figure 2. Release behavior of allyl sulfide from its α-CD inclusion complex spray-dried powders at different temperatures (: 100 °C, : 80 °C, : 60 °C). The solid lines were calculated by Equation (1). 0

-4

ln k10 ( )

inclusion complex spray-dried powders correlated well using Equation (1) with the Gaussian distribution of the ∆G values of the release rate constant. The release rate constant (k10) of the allyl sulfide/α-CD inclusion complex (5.95 × 10−2 day−1) was smaller than that of the β- and γ-CD (7.51 × 10−1 day−1) as shown in Table 2. This finding implies that the allyl sulfide/α-CD inclusion complex was much more stable than the β- and γ-CD complexes when stored under hot conditions. The flavor-release rate constant from a CD inclusion complex might depend on the type of CD, and the nature and polarity of the flavor molecules. Yoshii et al. [25] have reported that the (–)-menthol/γ-CD complex has a lower release rate constant than the (–)-menthol/β-CD complex. Li et al. [26] reported the release of AITC from the α-CD-AITC complex was lower than from its β-CD complex, and explained that AITC bound more strongly to α-CD as compared to β-CD because of the better matched molecular sizes of AITC and α-CD. In these results, the CD inclusion complexes could retain the flavor molecule for a short time at mild temperatures. However, in our research, allyl sulfide released from its CD complexes more slowly, which might be a result of the presence of MCT oil and the preparation method of the inclusion complexes.

851

-8

-12 2.6

2.7

2.8

2.9

3.0

3.1

1000/T (K-1)

Figure 3. Arrhenius plot of the release rate constants of allyl sulfide from the α-CD inclusion complexes.

852


Effect of relative humidity on the release behavior of allyl sulfide from the inclusion complex powders Figure 4 shows the release behavior of allyl sulfide from the three types of inclusion complexes at 50 °C and 6, 40, 54, and 73% RH. The IMC and particle diameters of the spray-dried α-, β-, and γ-CD powders were 8.9, 11.3, and 8.8 wt%, and 9.6, 15, and 12 μm, respectively. The release rate of the α-CD powder was slower than that of the β- and γ-CD powders under all RH conditions tested in this research. The release rates for all CDs had two steps in common, an initial rapid release followed by a slow release. This two-step release is best described by using a first-order release rate equation with a Gaussian distribution of the ∆G value of the release rate constant, as shown by the solid lines in Figure 4. Soottitantawat et al. [29] have concluded that RH is a major controlling factor for flavor release from spray-dried powders. Furthermore, Zhang et al. [31] have reported that low temperature and RH facilitated flavor retention.

The flavor release might take place via exchange of molecules in the cavity of CDs from the included flavor molecule to the water molecules [32]. Under high RH environments, the complex might adsorb large amounts of water on the outside of an inclusion complex, leading to the fast replacement of flavor and the collapse of the wall material [31,33], which might explain why the release rate increased with the increase in RH. Effect of IMC on the release behavior of allyl sulfide from the inclusion complex powders The IMC of the inclusion complex powders plays an important role in flavor release during storage. The effect of IMC on the allyl sulfide release rate was investigated by using α-, β-, and γ-CD powders with two different IMC (2.7 and 8.9 wt% for α-CD,1.8 and 11.3 wt% for β-CD, and 2.7 and 8.8 wt% for γ-CD) at 50 °C and 6, 40, 54, and 73% RH. As shown in Figure 5, the release rate of allyl sulfide from its inclusion complex powders with

Figure 4. Release behavior of allyl sulfide from its α-, β-, and γ-CD powders with initial moisture content of 8.9, 11.3, and 8.8 wt%, respectively, at different RH and 50 °C (, , , : α-CD with 6, 40, 54, 73% RH, ■, , , : β-CD with 6, 40, 54, 73% RH, ▲, △, , : γ-CD with 6, 40, 54, 73% RH, respectively). The solid lines were calculated by Equation (1).

Figure 5. Release behavior of allyl sulfide from its α-, β-, and γ-CD powders with initial moisture content of 2.7, 1.8, and 2.7 wt%, respectively, at different RH and 50 °C (, , , : α-CD with 6, 40, 54, 73% RH, ■, , , : β-CD with 6, 40, 54, 73% RH, ▲, △, , : γ-CD with 6, 40, 54, 73% RH, respectively). The solid lines were calculated by Equation (1).


3

ln k10 ( )

0 -3 -6 -9 -12

0

25

50

75

100

RH (%)

Figure 6. Relationship between RH and ln k10 at different IMC (, : α-CD powders with 8.9 and 2.7 wt% IMC, ■, : β-CD powders with 11.3 and 1.8 wt% IMC, and ▲, △: γ-CD powders with 8.8 and 2.7 wt% IMC; — IMC 8.9, 11.3, and 8.8 wt% of α-, β-, and γ-CD, respectively; ┄ IMC 2.7, 1.8, and 2.7 wt% of α-, β-, and γ-CD, respectively).

an IMC of 2.7 wt% was slow for all RH, whereas that from powders with an IMC of 8.9 wt% was higher at 73, 54, and 40% RH (Figure 4). With β- and γ-CD inclusion complex powders at the two different IMC values, the release rate of allyl sulfide was significantly affected by the increase in RH (Figures 4 and 5). The solid lines in Figures 4 and 5 were calculated using Equation (1). The standard deviation (σ) of the ∆G distributions was 3, 3.9, and 2.3 for α-, β-, and γ-CD, respectively. Plotting ln k10 vs. RH showed the dependence of the release rate constant k10 on the RH with the two IMC of the CD inclusion complex powders (Figure 6). The logarithm of the release rate constants for all CDs was linearly proportional to RH. According to these results, the release rate constant of the α-CD powders with an IMC of 2.7 wt% was not affected by the changes in RH, whereas all the other powders, with respect to IMC, were significantly affected by the changes in RH. With an IMC of 2.7 wt%, the release rate of the α-CD powder remained unchanged at different RH. In contrast, that of the α-CD powder with 8.9 wt% of IMC changed significantly when RH differed. As shown in Figure 6, the relationship between RH and ln k10 was considerably different, so it could be concluded that RH strongly affects the release of allyl sulfide from β-CD powders with two different IMC, and the IMC of the β-CD powder has a slight effect on the release rate of allyl sulfide at different RH. On the other hand, IMC has nearly no effect on the release rate of γ-CD powders at different RH. Therefore, the guest molecule was more stable for the α-CD inclusion complex without moisture adsorption, as compared to the β- and γ-CD inclusion complex powders. Tanada et al. [34] found that water molecules was adsorbed at the hydroxyl groups located on the surface and/or intermolecular spaces of α-CD, and at the hydroxyl groups on surface and in the cavity of β-CD.

853

In contract, the water molecules were only adsorbed in the cavity of γ-CD. The release of flavor from the CD cavity might involve an exchange reaction between the encapsulated flavor and water molecules. In the α-CD inclusion complex at a low IMC, this exchange reaction may not occur at 2.7 wt% because the water molecules would not be adsorbed in the cavity of α-CD. In our previous work [19], we demonstrated that the spray-dried powder of α-CD was in an amorphous state. The moisture adsorption rate for amorphous α-CD powder might be significantly affected by the IMC and this might affect the flavor release rate. Moreover, the MCT oil may inhibit of water adsorption because a part of the MCT oil in powder form exists as an oil-droplet without the formation of an inclusion complex. Cevallos et al. [35] reported that the inclusion complexes thymol-β-CD and cinnamaldehyde-β-CD remained stable up to 75% RH at 25 °C during long storage, and the release of the guests from the β-CD inclusion complexes was detected at 84% RH, at which the water adsorption rate increased. Fang et al. [36] reported that when limonene was encapsulated by β-CD, the apolar cavity was preferentially occupied by hydrophobic oil molecules, and the surface polar site bonds were associated with water. When water penetrated the encapsulating matrix, the system turned unfavorable for the hydrophobic guest, and the guest moved to a lower water content environment [37]. In the case of the β- and γ-CD inclusion complex powders prepared in our research, the bigger inner cavity might lead to a quick replacement of allyl sulfide by water molecules that were adsorbed at the surface of the β- and γ-CD. In addition, the solubility of γ-CD (23.2/100 g H2O) is more than 10-times higher than that of β-CD (1.85/100 g H2O) and 2-times higher than that of α-CD (14.5/100 g H2O). Nakai et al. [38] investigated the properties of crystal water of α-, β-, and γ-CD, and showed that α-CD in its hydrate form was stable at RH levels above 11%. Consequently, the adsorption rate of water might be slower than that of β- and γ-CD.

Conclusion The release rates of allyl sulfide from CDs depended markedly on the RH. The IMC of 2.7 wt% of α-CD did not affect the release rate of allyl sulfide from the α-CD powder. The release behavior of allyl sulfide from CDs inclusion complex powders correlated well with the first-order release rate equation with the Gaussian distribution of ∆G of the release rate constant.

Author contributions Thi Van Anh Nguyen and Hidefumi Yoshii designed this study and discussed the results. Thi Van Anh Nguyen carried out the experiments and wrote the manuscript. Hidefumi Yoshii revised the manuscript. Both the authors have read and approved the final manuscript.

854


Acknowledgement Thi Van Anh Nguyen is grateful to the Vietnam Ministry of Education and Training for financial support through the Vietnamese Government Scholarship (911 Project) during the course of study.

Disclosure statement No potential conflict of interest was reported by the authors.

Funding This work was supported by Grant-in-Aid for Scientific Research (C) (No. 15K07455) for Hidefumi Yoshii.

References [1] Szejtli J. (1988), Cyclodextrin. In Davis JED (Eds), Cyclodextrin technology. Dordrecht: Kluwer Academic. p. 2–15. [2] Bhandari BR, D’Arcy BR, Le TBL. Lemon oil to α-cyclodextrin ratio effect on the inclusion efficiency of α-cyclodextrin and the retention of oil volatiles in the complex. J Agric Food Chem. 1998;46:1494–1499. [3] Yoshii H, Furuta T, Okita E, et al. The increased effect of kneading on the formation of inclusion complexes between d-limonene and β-cyclodextrin at low water content. Biosci Biotechnol Biochem. 1998;62:464–468. [4] Shiga H, Yoshii H, Nishiyama T, et al. Flavor encapsulation and release characteristics of spray-dried powders by the blended encapsulant of cyclodextrin and gum Arabic. Dry Technol. 2001;19(7):1385–1395. [5] Jafari SM, Assadpoor E, He Y, et al. Encapsulation efficiency of food flavors and oils during spray drying. Dry Technol. 2008;26(7):816–835. [6] Amagase H. Clarifying the real bioactive constituents of garlic. J Nutr. 2006;136:716S–725S. [7] Strickley RG. Solubilizing excipients in oral and injectable formulations. Pharmaceut Res. 2004;21:201–230. [8] Turchiuli C, Jimenez Munguia MT, Hernandez Sanchez M, et al. Use of different supports for oil encapsulation in powder by spray drying. Powder Technol. 2014;255:103–108. [9] Yang Y, McClements DJ. Encapsulation of vitamin E in edible emulsions fabricated using a natural surfactant. Food Hydrocolloids. 2013;30:712–720. [10] Madene A, Jacquot M, Scher J, et al. Flavour encapsulation and controlled release – a review. Int J Food Sci Technol. 2006;41:1–21. [11] Kant A, Linforth RST, Hort J, et al. Effect of β-cyclodextrin on aroma release and release perception. J Agric Food Chem. 2004;52:2028–2035. [12] Shiga H, Furuta T, Yoshii H, et al. Release characteristic of allyl isothiocyanate encapsulated in cyclodextrins. In Proceedings of the 10th International Cyclodextrin Symposium; Ann Arbor, MI; 2000. p. 553–558. [13] Chin ST, Nazimah SHA, Quek SY, et al. Effect of thermal processing and storage condition on flavor stability of spray-dried durian powder. LWT-Food Sci Technol. 2010;43:856–861. [14] Capozzi LC, Bazzano M, Cavallero MC, et al. Polymeric supports for controlled release of ethylene for food industry. Int Polym Proc. 2016;31(5):570–576. [15] Shiga H, Yoshii H, Taguchi R, et al. Release characteristic of flavor from spray-dried powder in boiling water

and during rice cooking. Biosci Biotechnol Biochem. 2003;67:426–428. [16] Wang T, Li B, Si H, et al. Release characterisation and antibacterial activity of solid state eugenol/βcyclodextrin inclusion complex. J Incl Phenom Macro. 2011;71:207–213. [17] Yoshii H, Furuta T, Fujiwara M, et al. Oxidation stability of powdery ethyl eicosapentaenoate included in cyclodextrin and polysaccharide/cyclodextrin mixtures. Jpn J Food Eng. 2003;4:25–30. [18] Kawamura Y, Nakanishi K, Matsuno R, et al. Stability of immobilized & #xF061;-chymotrypsin. Biotechnol Bioeng. 1981;23(6):1219–1236. [19] Ishido E, Hakamata K, Minemoto Y, et al. Oxidation process of linoleic acid encapsulated with a polysaccharide by spray-drying. Food Sci Technol Res. 2002;8:85–88. [20] Wang J, Cao Y, Sun B, et al. Physicochemical and release characterisation of garlic oil-β-cyclodextrin inclusion complexes. Food Chem. 2011;127:1680–1685. [21] Ho TM, Howes T, Bhandari BR. Dehydration of CO2-αcyclodextrin complex powder by desiccant adsorption method and its release properties. J Microencapsul. 2016;33(8):763–772. [22] Nguyen TVA, Yoshii H. Encapsulation of allyl sulfide with middle-chain triglyceride oil and cyclodextrin by spray drying. Jpn J Food Eng. 2017;18:35–42. [23] Soottitantawat A, Yoshii H, Furuta T, et al. Effect of water activity on the release characteristics and oxidative stability of d-limonene encapsulated by spray drying. J Agric Food Chem. 2004;52(5):1269–1272. [24] Greenspan L. Humidity fixed point of binary saturated aqueous solution. J Res Natl Bur Stand. 1977;81A(1):89– 96. [25] Yoshii H, Sakane A, Kawamura A, et al. Release kinetics of (–)-menthol from chewing gum. J Incl Phenom Macro. 2007;57:591–596. [26] Li X, Jin Z, Wang J. Complexation of allyl isothiocyanate by α- and β-cyclodextrin and its controlled release characteristics. Food Chem. 2007;103:461–466. [27] Reineccius TA, Reineccius GA, Peppard TL. Flavor release from cyclodextrin complexes: comparison of alpha, beta, and gamma types. J Food Sci. 2003;68:1234– 1239. [28] Hadaruga DI, Hadaruga NG, Rivis A, et al. Thermal and oxidative stability of the Allium sativum L. bioactive compounds/α- and β-cyclodextrin nanoparticles. Rev Chim. 2007;58(11):1009–1015. [29] Soottitantawat A, Takayama K, Okamura K, et al. Microencapsulation of l-menthol by spray drying and its release characteristics. Innov Food Sci Emerg. 2005;6:163–170. [30] Zhu G, Xiao Z, Zhou R, et al. Kinetics and release characteristics of menthyl acetate from its β-cyclodextrin inclusion complex by thermogravimetric analysis. J Incl Phenom Macro. 2016;84:219–224. [31] Zhang Y, Zhou Y, Cao S, et al. Preparation, release and physicochemical characterization of ethyl butyrate and hexanal inclusion complexes with cyclodextrin. J Microencapsul. 2015;32:711–718. [32] Rehmann L, Yoshii H, Furuta T. Characteristics of modified β-cyclodextrin bound to cellulose powder. Starch. 2003;55(7):313–318. [33] Yoshii H, Soottitantawat A, Liu XD, et al. Flavor release from spray-dried maltodextrin/ gum arabic or soy matrices as a function of storage relative humidity. Innov Food Sci Emerg. 2001;2:55–61.


[34] Tanada S, Nakamura T, Kawasaki N, et al. Adsorption behaviour of water molecules onto α-, β-, and γ-cyclodextrins and branched α-cyclodextrins. J Colloid Interface Sci. 1996;181:326–330. [35] Cevallos PAP, Buera MP, Elizalde BE. Encapsulation of cinnamon and thyme essential oil components (cinnamaldehyde and thymol) in β-cyclodextrin: effect of interactions with water on complex stability. J Food Eng. 2010;99(1):70–75. [36] Fang Z, Comino PR, Bhandari BR. Effect of encapsulation of D-limonene on the moisture

855

adsorption property of β-cyclodextrin. LWT-Food Sci Technol. 2013;51:164–169. [37] Ayala-Zavala JF, Soto-Valdez H, González-León A, et al. Microencapsulation of cinnamon leaf (Cinnamomum zeylanicum) and garlic (Allium sativum) oil in β-cyclodextrin. J Incl Phenom Macrocycl Chem. 2007;60:359–368. [38] Nakai Y, Yamamoto K, Terada K, et al. Properties of crystal water of α-, β-, and γ-cyclodextrin. Chem Pharm Bull. 1986;34:2178–2182.