Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ldrt20 Drying Technology An International Journal ISSN: 0737-39
Trang 1Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ldrt20
Drying Technology
An International Journal
ISSN: 0737-3937 (Print) 1532-2300 (Online) Journal homepage: http://www.tandfonline.com/loi/ldrt20
Kinetic Analysis and Evaluation of Controlled
Release of D-Limonene Encapsulated in
Spray-Dried Cyclodextrin Powder under Linearly Ramped Humidity
Chisho Yamamoto , Tze Loon Neoh , Hirokazu Honbou , Hidefumi Yoshii & Takeshi Furuta
To cite this article: Chisho Yamamoto , Tze Loon Neoh , Hirokazu Honbou , Hidefumi Yoshii
& Takeshi Furuta (2012) Kinetic Analysis and Evaluation of Controlled Release of D-Limonene Encapsulated in Spray-Dried Cyclodextrin Powder under Linearly Ramped Humidity, Drying
Technology, 30:11-12, 1283-1291, DOI: 10.1080/07373937.2012.681089
To link to this article: https://doi.org/10.1080/07373937.2012.681089
Published online: 17 Aug 2012
Submit your article to this journal
Article views: 172
View related articles
Citing articles: 5 View citing articles
Trang 2Kinetic Analysis and Evaluation of Controlled Release
Powder under Linearly Ramped Humidity
Chisho Yamamoto,1,2 Tze Loon Neoh,2Hirokazu Honbou,2 Hidefumi Yoshii,2 and
Takeshi Furuta3
1
Tottori Institute of Industrial Technology, Tottori, Japan
2
Department of Applied Biological Science, Kagawa University, Kagawa, Japan
3
Department of Chemistry and Biotechnology, Tottori University, Tottori, Japan
A linearly ramped humidity system coupled with automatic
sampling gas chromatography has been developed The apparatus
was adopted to measure the dynamic release flux of D-limonene
included in spray-dried b-cyclodextrin (b-CD) and the
polymer-coated counterparts under linearly ramped humidity (0.375%/min)
at constant temperature (50C) The release flux profiles ofD
-limon-ene were analyzed using a mathematical model based on an
extended Arrhenius equation The activation energy ofD-limonene
release flux from b-CD was 278 (kJ/mol), which was the highest
among the powders and around twice as high as our previous results
obtained by the conventional static method
Keywords Dynamic flavor release; Encapsulation by
cyclodex-trin; Humidity ramping method; Spray drying
INTRODUCTION
Cyclodextrins (CDs) are doughnut-shaped cyclic
oligosaccharides with interior cavities capable of forming
specific inclusion complexes with many organic
com-pounds CDs are often used as encapsulants of
hydro-phobic flavors, a method called molecular encapsulation
because the flavors are encapsulated in the molecular
cavi-ties of CDs After a century of continuous research and
development, CDs have gained certain recognition in
vari-ous fields.[1]Their applications are mainly intended for the
entrapment of smaller molecules, catalysis through
encap-sulation, and as potential molecular transport devices In
food-related applications, flavor compounds are being
encapsulated into CDs for better retention and protection
from various possible means of deterioration, as well as
for controlled delivery.[1–4] In addition to food flavors,
CDs are used for encapsulation of antibacterial
com-pounds such as allyl isothiocyanate and hinokitiol
(b-thujapilicin) for potential applications in paper and cloth.[5–8]
For the manufacturing of flavor powders, high flavor retention is an important quality factor but the controlled release characteristics of the encapsulated flavors from the powders are equally important in terms of storage stab-ility and application A number of studies on flavor release have been conducted, and it has been reported that the fla-vor releases were closely related to the relative humidity (RH) of the storage environment.[1,9–13]In most of the stu-dies on flavor release from powders, the rate of release was measured with a thin packed layer of the powders under constant temperature and humidity conditions (static method) The static method of characterizing flavor release
is extremely time intensive, and it takes a few weeks to obtain the full release profiles of flavors from the pow-ders.[1,12] In addition, the method does not provide any additional observations triggered by the rapid structural evolution such as aggregation and surface solubilization
of the powder particles, which is often reflected by abrupt changes in the release flux
Dronen and Reineccius[14] and Mortenson and Reineccius[15,16] applied a proton transfer reaction mass spectrometer (PTR-MS) coupled to a dynamic vapor sorp-tion (DVS) instrument as a rapid method of analysis to measure the release time course of flavors from spray-dried powders They obtained dynamic release profiles of flavors but did not report the establishment of the kinetics of the release reactions Mateus et al.[17] measured the release profiles of volatile organic compounds from roasted and ground coffee beans using PTR-MS and analyzed the release profiles with the empirical models of diffusion and Weibull’s equation.[12] Bohn et al.[18] also applied the method to an amorphous sucrose–based glassy matrix encapsulating cherry flavor and correlated the flavor rele-ase with the glass transition temperature of the matrix.[19] Although it is well known that the release of flavors
Correspondence: Takeshi Furuta, Department of Chemistry
and Biotechnology, Tottori University, 4-101, Koyama, Tottori
680-8552, Japan; E-mail: furuta@sun.ocn.ne.jp
ISSN: 0737-3937 print=1532-2300 online
DOI: 10.1080/07373937.2012.681089
1283
Trang 3through the matrices of the solid state is closely related to
environmental humidity, few studies have been conducted
to develop a kinetic model of flavor release as a function
of humidity Normand et al.[20] investigated the release
profile ofD-limonene from empty yeast cells by monitoring
the weight change with DVS at various humidities They
correlated the release flux with a modified Arrhenius
equa-tion Zhao et al.[21]and Li et al.[22]analyzed solid-state drug
stabilities by applying an analogous Arrhenius equation
by simultaneous varying temperature and humidity In
addition to the DVS method, Neoh and coworkers[23,24]
studied the release mechanism of organic compounds
inclu-ded in CDs using thermogravimetry However, the studies
were not conducted under humid conditions
Almost all of the studies on flavor release have been
car-ried out under constant humidity conditions However,
more informative data can be acquired quickly by applying
variable humidity conditions In this study, a rapid
evalu-ation of the release of D-limonene from spray-dried
b-CD=D-limonene complex powders was performed under
linearly ramped humidity at constant temperature using a
home-built DVS system Real-time release fluxes (release
rates) ofD-limonene were monitored by autosampling gas
chromatography The release flux profiles under ramping
humidity were modeled by applying an extended Arrhenius
equation The kinetic parameters such as activation energy
and frequency factor were estimated The impacts of
poly-mer coating agents on the release mechanism were also
investigated kinetically
MATERIALS AND METHODS
Materials
Cavamax W7 (standard grade b-cyclodextrin, b-CD,
Wacker Chemie AG, Stuttgart, Germany) was obtained
from Cyclochem Co., Ltd (Kobe, Japan) Better Sol, sup-plied by Seishin Enterprise Co., Ltd (Tokyo, Japan), which is a soap-free colloidal emulsion made of polyethyl-ene resin (PE) and various cross-linkers (solids content of 20%) was used as a coating agent Sumikaflex, a water-based emulsion composed of ethylene-vinyl acetate (EVA) copolymer resins, which was also used as a coating agent, was purchased from Sumika Chemtex Co., Ltd (Tokyo, Japan) as a 50% emulsion solution The ethylene unit con-tent of EVA was approximately 10%.D-Limonene was pur-chased from Nacalai Tesque, Inc (Kyoto, Japan) Distilled water was used throughout the entire study Unless other-wise stated, all other chemicals used in this study were pur-chased from Wako Pure Chemical Industries, Ltd (Osaka, Japan) and were of analytical reagent grade
Preparation of Complex Powders by Spray Drying
D-Limonene was selected as a model flavor (guest com-pound) Three hundred grams of 33.9 wt% b-CD solution was prepared in a 500-mL laboratory bottle One and a half molar times of D-limonene relative to b-CD were added The total concentrations of b-CD andD-limonene were set constant at 40 wt% in all treatments The mixture solution was stirred with a magnetic stirrer for 24 h After complexation equilibrium, the polymer coating agent (Better Sol or Sumikaflex) was added to the mixture sol-ution at 0 to 12 wt% Because Better Sol and Sumikaflex contain polymer resins, they create a polymer coating after spray drying The feed liquid was homogenized with a Polytron homogenizer (model PT-6100, Kinematica Inc., Bohemia, NY) for 3 min at a rotor speed of 8,000 rpm The compositions of the feed liquids were listed in Table 1a Spray drying was performed with an Ohkawara-L8 spray dryer (Ohkawara Kakouki Co., Ltd., Yokohama,
TABLE 1 Composition of feed liquid, properties of spray-dried powder, and kinetic parameters of release
Spray-dried powders
(a) Feed liquid composition b-CD (%) 33.9 33.9 33.9 33.9
(b) Properties of spray-dried
powder
D-Limonene content (g=g-powder) 0.087 0.091 0.088 0.079
(c) Kinetic parameters of
release ofD-limonene
Uncoated: spray-dried b-CD=D-limonene complex powder: PE1: PE (6%)-coated spray-dried complex powder: PE2: PE (12%)-coated spray-dried complex powder: EVA: EVA (4.8%)-coated spray-dried complex powder
Trang 4Japan) The size of the spray-drying tower has been
described elsewhere.[25] The feed mixture was fed at
30 mL=min to the rotary disk atomizer and atomized at a
speed of 10,000 rpm The drying medium (air) of 200C
was supplied at 110 kg=h with an outlet air temperature
ranging between 145 and 150C The spray-dried powders
were collected in the cyclone and put in a hermetically
sealed container and stored at30C until use The
spray-dried b-CD=D-limonene complex powder and the PE
(6%)-coated, PE (12%)-coated, and EVA (4.8%)-coated
spray-dried complex powders will be referred to
through-out the remainder of this article as uncoated, PE1, PE2,
and EVA, respectively
Quantification ofD-Limonene in Spray-Dried Powders
Solvent extraction was employed for quantification of
D-limonene in the spray-dried powders The spray-dried
complex powder (0.1 g) was added with 4 mL of water
and 1 mL of chloroform into a glass vial, which was tightly
screw-capped, and vigorously vortexed for 1 min The
chloroform contained cyclohexanone at a concentration
of 1 mL=mL chloroform, serving as an internal standard
in gas chromatographic determination of D-limonene To
extract theD-limonene into the organic solvent, the mixture
was heated at 90C for 30 min with several periods of
inter-mittent vortexing After cooling of the extracted samples
to room temperature, a 10-min centrifugation at
3,000 rpm was performed The samples were extracted in
triplicate for quantification by gas chromatography
A 1-mL aliquot of the supernatant was injected in
dupli-cate into a GC-14A gas chromatograph (Shimadzu Corp.,
Kyoto, Japan) Chromatographic separation was
per-formed using a 20% PEG-20 M-packed (Shinwa Chemical
Industries, Ltd., Kyoto, Japan) glass column (3.2 mm i.d
2 m-length) with nitrogen as the carrier gas (0.07 MPa)
D-Limonene was quantified using a flame ionization
detec-tor (FID) The temperatures of the injection port, oven,
and detector were set constant at 140, 130, and 230C,
respectively To obtain a calibration value for
quantifi-cation ofD-limonene, two bottles of standard solution of
differentD-limonene contents (1.6 and 3.2 mL=mL ofD
-lim-onene) containing 1 mL=mL cyclohexanone in chloroform
were prepared A 1-mL aliquot of each standard solution
was injected into the gas chromatograph twice, and the
average peak areas were used as a calibration value to
calculate the amount ofD-limonene
Size and Structure of Spray-Dried Powder Particles
The size distributions of spray-dried particles were
measured with a laser scattering particle size analyzer
(SALD-7100, Shimadzu Corp.) installed with a batch
sam-ple cell.[26] The powder (100 mg) was dispersed in 2 mL of
2-methyl-1-propanol, in which the powder was not miscible
nor did it agglomerate In addition, 2-methyl-1-propanol
does not affect the particle size by shrinkage or expansion and has high dispersibility and wettability About 50 mL of each dispersion was introduced into the batch sample cell containing 6.5 mL of 2-methyl-1-propanol The volume-averaged diameter, D43, was used as the mean diameter for all measurements Each sample was analyzed in duplicate
A scanning electron microscope (SEM; JSM 6060, JEOL Co., Ltd., Tokyo, Japan) was used to observe the microstructure of the spray-dried particles The powder particles were mounted onto the SEM stubs using double-sided adhesive carbon tape (Nisshin EM Co., Ltd., Tokyo, Japan) In order to examine the inner structure, the micro-capsules (attached to the stub) were fractured by adhering
a second piece of adhesive tape on top of the samples and then quickly ripping them apart.[27]The specimen stub was subsequently coated with Pt-Pd using an MSP-1S mag-netron sputter coater (Vacuum Device Inc., Tokyo, Japan) The sputter-coated sample was then analyzed using the SEM operated at 2.0 kV
Measurement of Release Flux ofD-Limonene from Inclusion Complex Powders
A home-built DVS system coupled with an autosam-pling gas chromatograph was adopted to conduct the rapid evaluation of the release of D-limonene from spray-dried
CD powders under linearly ramped humidity at constant temperature The release equipment consisted of two parts: the DVS system, which humidifies the nitrogen flow
at the prescribed rate, and the quantification system of the release flux ofD-limonene from the powders under varying RH
Nitrogen was used as a carrier gas based on the prelimi-nary finding that revealed that D-limonene vapor reacted immediately with oxygen in air, producing several gaseous oxide compounds Nitrogen was vapor-saturated by bubbling the nitrogen gas through distilled water in a 500-mL laboratory bottle as shown in Fig 1 A waterproof heater was wound around the bubbling bottle and immersed in a water bath The temperatures of the water bath and the water in the bubbling bottle were controlled using a programmable temperature controller (DSSP93, Shimaden, Tokyo, Japan) through a thermocouple in the water bath The vapor-saturated nitrogen was passed through a mist trap before being flowed into an air bath through a stainless steel pipe wound with a guard heater The distilled water in the bubbling bottle was stirred with
a magnetic stirrer to ensure thorough dispersion and suf-ficient residence time for vapor saturation in the bubbles The nitrogen flow rate was 100 mL=min In order to achieve an RH profile ramping linearly from 10 to 100% within a prescribed time, the water vapor pressure– temperature relation was divided and linearized into seven sections over the temperature range between 10 and 60C,
Trang 5and data sets of time and temperature pairs were input into
the temperature controller
The vapor-saturated nitrogen flowing into the air bath
was equilibrated at the constant temperature (40, 50, or
60C) of the bath by passing through a spiral copper tube
(3.2 mm i.d. 5 m length) at the inlet of the bath The RH
of the nitrogen gas was estimated as p=ps, where p is the
satu-rated water vapor pressure at the current temperature of the
bubbling bottle and psis the saturated water vapor pressure
at the air bath temperature The humidity-conditioned
nitro-gen was flowed through the glass release vessel (16 mm
i.d. 80 mm height) at the bottom of which the aluminum
pan (13 mm i.d. 1 mm depth) packed with the sample
pow-der was set, as shown in Fig 1 TheD-limonene released from
the sample powder was carried away by the nitrogen stream
into the gas chromatograph for quantification The real-time
release flux of D-limonene per unit mass of the sample
powder (release flux), F (mg=s cm2 g-powder), was
calcu-lated by multiplying the concentration ofD-limonene in the
nitrogen stream Cg (mg=mL) by the nitrogen flow rate V
(mL=s) and dividing by the mass of the powder, m (g) and
the surface area, A (cm2), of the powder:
F ¼ V Cg=ðA mÞ ð1Þ The concentration ofD-limonene in the nitrogen stream,
Cg, was measured in real time with a gas chromatograph
(GC-14B, Shimadzu) Sampling of the nitrogen stream and
injection were automated by a timer-operated switching
valve (Valco A6-G6 W, Valco Instruments Co Inc., TX)
In sampling mode, each port of the valve was connected as
shown by the solid lines in Fig 1 The effluent nitrogen from
the release vessel was flowed through a 5.0-mL sampling
loop and exhausted In injection mode, the valve was
switched by compressed air and the connections between
ports were changed to that shown by the dotted lines in Fig 1 The carrier gas from the gas chromatograph swept out the sampled effluent nitrogen from the sampling loop into a PEG-20 M-packed glass column for separation and finally for quantification of current Cgin an FID Two stan-dard solutions of different D-limonene contents in chloro-form (1.0 and 2.0 mL=mL of D-limonene) were prepared
An absolute calibration curve was made by injecting a 1-mL aliquot of each standard solution into the gas chroma-tograph twice and averaging the peak areas ofD-limonene
Cgwas calculated with the calibration curve The switching valve alternated between a 4-min sampling mode and a 1-min injection mode for the entire release experiment because the whole cycle of 5 min was necessitated by the limi-tation of the chromatographic separation The humidity of the nitrogen stream was continuously monitored with a humidity sensor (HMP233, Vaisala, Helsinki, Finland) connected to the effluent port of the valve
RESULTS AND DISCUSSION Changes of RH in Nitrogen Stream under Linearly Ramping Program
In this study, the RH of the nitrogen stream was chan-ged in a linearly ramping manner For this purpose, the temperature of the bubbling bottle was regulated with the programmable temperature controller by segmenting the water vapor pressure–temperature relationship into seven linear sections Figure 2 indicates the time course
of RH under the ramping program from 10 to 100% for
4 h; the rate of increase (ramping rate) was 0.375%=min The humidity was monitored at the exit of the release apparatus as mentioned in the previous section The humidity of the nitrogen stream satisfactorily increased lin-early with time, thus warranting the performance of the home-built DVS system The time course of the water FIG 1 Schematic diagram of the home-built DVS system coupled with an autosampling gas chromatograph.
Trang 6temperature in the bubbling bottle is also illustrated in
Fig 2 The temperature profile is not linear but curved
upwards against time
Content ofD-Limonene and Morphologies of
Spray-Dried Particles
D-Limonene contents and particle sizes of the spray-dried
powders are listed in Table 1b All four powders contained
approximately 0.8 molar ratio ofD-limonene with respect to
b-CD regardless of the type of polymer coating agent in use
In contrast, the powder particle size varied drastically with
both the amount and type of coating agent The addition of
the coating agents may have increased the viscosity of the
feed liquids, which is one of the key factors determining
the diameter of the spray droplet; the more viscous the feed
liquid is, the larger the atomized droplet size is.[25]
The external and internal morphologies of the
spray-dried particles are illustrated in Fig 3 The uncoated
complex powder (Fig 3a) was very fine crystal particles of
several micrometers, and the coated complex powders
ran-ged from a few to several tens of micrometers As seen in
Figs 3b–3d, the coated particles were spherical in shape
and their surfaces were composed of the fine complex
parti-cles bound by the coating agents of PE and EVA Many
pores appeared between the fine complex particles on the
powder particle surfaces The surface of the EVA-coated
powder particles was rougher than that of the PE-coated
ones This might be attributed to the lower hydrophobicity
of EVA compared to that of PE The internal structures of
the coated powder particles are shown in Figs 3e–3g Fine
complex particles packed in a random manner on the inside
of the coated complex powder particles, forming relatively
rougher and more porous structures compared to the
sur-faces A clear boundary discerning the outer crust from
the inner core (randomly packed fine complex particles)
was vividly observable in the powder particles coated with
PE (Figs 3e and 3f), whereas in EVA-coated particles, such
a well-defined boundary was not clearly noticeable (Fig 3g) These differences in particle surface structure may have cru-cial impacts on the release characteristics ofD-limonene
Release Characteristics ofD-Limonene from Spray-Dried Powders
The release flux of D-limonene from the spray-dried powders in the solid state was measured under the RH ramping rate of 0.375%=min at 40, 50, and 60C with the DVS-GC system shown in Fig 1 As shown in Fig 4, the release flux, F, for the uncoated spray-dried complex pow-der was nearly zero up to 70% RH and increased monoto-nously with a further increase in RH Particularly at 60C, the release flux, F, grew exponentially with the increase in
RH above 80% As for the PE-coated powders, the release flux, F, was markedly decreased compared to that of the uncoated complex powder Higher amounts of the coating agent inhibited the release ofD-limonene to a more remark-able extent In the case of PE2-coated powder (Fig 4c), the release of D-limonene was almost negligible at RH < 90%
FIG 2 Linearly ramped relative humidity in N 2 stream (.) at the
corre-sponding temperatures of the bubbling water () The humidity ramping
rate was set at 0.375% RH=min.
FIG 3 SEM micrographs of spray-dried particles illustrating the exter-nal and interexter-nal morphologies of the (a) uncoated complex powder and the complex powders coated with (b), (e) PE (6.0%, PE1); (c), (f) PE (12%, PE2); and (d), (g) EVA.
Trang 7for all of the temperatures used Because PE is a
hydro-phobic compound, the water vapor in the nitrogen stream
would be prevented from adsorbing into the particles by
the PE film covering the particle surface Consequently,
D-limonene included in b-CD could hardly be released
from the molecular cavity In contrast, D-limonene began
to release from the EVA-coated particles at an earlier stage
before 70% RH The flux, F, increased exponentially with
increasing RH and exhibited higher values at the same
RH compared to other particles, particularly at 40 and
50C This may be directly ascribable to the hydrophilicity
of EVA EVA increasingly adsorbed moisture at higher
RH and might contribute to triggering the massive release
ofD-limonene At 60C, however, the release flux ceased to increase further at RH higher than 80% but assumed a release profile similar to that of 50C This phenomenon may possibly be caused by the collapse of the surface struc-ture of the particles resulting from the melting of the fine complex particles at higher RH Figure 5 shows the sur-faces of particles coated with PE1 and EVA that had been exposed to the nitrogen stream of 80% RH for 120 min For the particles coated with PE, the fine complex particles
FIG 4 Release profile of D -limonene under linearly ramped humidity (0.375% RH=min) at 40 C (4), 50 C (), and 60 C ( & ) from the (a) uncoated complex powder and the complex powders coated with (b) PE1, (c) PE2, and (d) EVA.
FIG 5 Changes of the surface structures of PE1- and EVA-coated complex powder particles exposed at 80% RH for 2 h.
Trang 8at the powder particle surface agglomerated to a lesser
extent after humidification in comparison to the
EVA-coated particles in which the surfaces were visually
dis-solved, causing structural collapse and filling up the pores
The reduction of the ascent rate of F at 60C and
RH > 80% observed in Fig 4d may be due to the inhibition
of diffusion ofD-limonene from the inside of the complex
powder particle to its surface
Kinetic Analysis of Release Flux ofD-Limonene at
Various Humidity and Temperatures
On the basis of an extended Arrhenius equation
pro-posed by Normand et al.,[20]in which the release ofD
-lim-onene was assumed to be a function of both temperature
and RH, a kinetic analysis of the release flux was
conduc-ted to estimate the activation energies and frequency
fac-tors of the release reactions The extended Arrhenius
equation is an empirical equation expressed as:
F ¼ A exp E
RT
expðB uÞ ð2Þ
where T, u, E, and R are the temperature, relative
humi-dity, activation energy, and gas constant, respectively A
is a constant, but B is a function of T as explained below
Equation (2) has also been applied by Zhao et al.[21]and Li
et al.[22] to analyze the solid-state drug stabilities Taking
the natural logarithm of both sides of Eq (2), one can
obtain the following equation:
ln F ¼ ln A E
RTþ B u ð3Þ Equation (3) describes that at a constant temperature, ln F
is linearly proportional to the relative humidity, u To
examine the validity of Eq (3), a semi-logarithmic plot of
F against u (ln F vs u) was constructed as shown in
Fig 6, using the release data in Fig 4a At constant
tem-perature, ln F can be clearly expressed by a linear equation
of u in a higher range of u, and B can be obtained from the
slopes of the plots However, the slopes vary at different
temperatures, implying that B in Eq (2) is not constant
but a function of temperature and is assumed to be
B¼B
0
where B0and C are constants If the y-intercept of the line
in Fig 6 were denoted as ln F0, the following equation
could be derived from Eq (3) at temperature T:
ln F0¼ ln A E
The activation energy of the release flux, E, could be calcu-lated from the slope of the plot of ln F0 against 1=T Figure 7 illustrates both plots for ln F0and B against the inverse of T Although there were a few variations, ln F0
and B still correlated well with 1=T The estimates of the four parameters, E, A, B0, and C, are listed in Table 1c The activation energy of release flux for the uncoated com-plex powder was the highest among the powders Because there are few reported studies concerning the release of molecularly encapsulated flavors from CDs in the solid-state condition, appropriate comparison of the results was difficult Furuta et al.[1] investigated the release of
D-limonene from b-CD at 50% RH by the static method and estimated the activation energy to be 123 kJ=mol Although the value seems much lower than the present results, it may be reasonable because the activation energy was obtained at a lower RH (50% RH) Because E in Eq (2) can be regarded as the activation energy of the release flux at zero relative humidity, the activation energies in Table 1c may be rational because much energy would be
FIG 6 Correlation of F as a function of relative humidity, u, at 40 C (4), 50 C (), and 60 C ( & ) for the uncoated spray-dried complex powder The humidity ramping rate was 0.375% RH=min.
FIG 7 Arrhenius plots of ln F 0 and B for the uncoated complex powder () and the complex powders coated with PE1 (4), PE2 ( 4 ), and EVA ( ).
Trang 9needed to dissociate the complex molecules between the
guest and host compounds under dry conditions.[1]
Normand et al.[20] conducted release experiments for
D-limonene enclosed in yeast cells under stepwise variation
of humidity and reported the activation energy of the
release flux to be 74 kJ=mol Zhao et al.[21]and Li et al.[22]
suggested that the activation energies of degradation were
76 kJ=mol for penicillin potassium and 93.5 kJ=mol for
aspirin under simultaneous variation of temperature and
humidity However, all of these studies were carried out
with physically encapsulatedD-limonene or intact
pharma-ceuticals BecauseD-limonene inclusion complex is bonded
in the molecular cavity of CD by Van der Waals forces and
a hydrophobic interaction, it is reasonable to consider that
more energy would be needed for the dissociation of the
complexed ingredient
Figure 8 shows the plot of ln A as a function of E A
sat-isfactory linear proportionality supported the
establish-ment of a compensation effect for the apparent kinetic
parameters Validation of the compensation effect may be
indicative of the fact that the release ofD-limonene in the
four samples possibly occurred via mechanisms of a
com-mon nature
As described earlier, because few studies have been
con-ducted to measure the dynamic release of encapsulated
fla-vors under varying humidity conditions and analyze it
kinetically, the validity of the present results such as the
activation energy of release could not be properly
evalu-ated The flavor release rate from powders may be
con-trolled mainly by molecular diffusion inside the powder
particle, which is strongly influenced by the matrix
struc-ture of the particle The dynamic release measurements
can respond sensitively to the structural changes such as
collapses and melting caused by the adsorption of the
moisture (Fig 4d) The subsequent interests of the study
lie in developing a mathematical model of the counterdiffu-sion of the flavor and moisture inside a particle, including the phase change of the matrix of the particle
CONCLUSION Real-time release fluxes of D-limonene included in spray-dried b-CD powder and their polymer-coated coun-terparts were measured using the home-built DVS system coupled with an autosampling gas chromatograph The release ofD-limonene from spray-dried b-CD and their cor-responding coated powders was investigated under linearly ramped humidity at constant temperature The release fluxes were nearly zero below 70% RH and increased monotonously with an increase in RH Higher amounts
of the coating agent inhibited the release to a more remark-able extent At constant temperature, the semi-logarithms
of the release fluxes correlated linearly with the higher range of relative humidity An extended Arrhenius equa-tion was applied and the kinetic parameters such as acti-vation energy and frequency factor were estimated The activation energy ofD-limonene release from the uncoated complex powder was highest and around twice as high as the previous results obtained by the static method REFERENCES
1 Furuta, T.; Soottitantawat, A.; Neoh, T.L.; Yoshii, H Effect of microencapsulation on food flavors and their release In Physico-chemical Aspects of Food Engineering and Processing; Devahastin, S., Ed.; CRC Press: New York, 2010; 3–40.
2 Szente, L.; Szejtli, J Molecular encapsulation of natural and synthetic coffee flavour with b-cyclodextrin Journal of Food Science 1986, 51(4), 1024–1027.
3 Reineccius, T.A.; Reineccius, G.A.; Peppard, T.L Encapsulation of flavors using cyclodextrins: comparison of flavor retention in alpha, beta, and gamma types Journal of Food Science 2002, 67(9), 3271– 3279.
4 Reineccius, T.A.; Reineccius, G.A.; Peppard, T.L Flavor release from cyclodextrin complexes: comparison of alpha, beta, and gamma types Journal of Food Science 2003, 68(4), 1234–1239.
5 Rehmann, L.; Yoshii, H.; Furuta, T Characteristics of modified b-cyclodextrin bound to cellulose powder Starch=Starke 2003, 55(7), 313–318.
6 Furuta, T.; Kusuya, Y.; Neoh, T.L.; Rehmann, L.; Beak, S.H.; Yoshii, H Inclusion and release of hinokitiol into=from MCT-b-CD fixed on Japanese washi paper Journal of Inclusion Phenomena and Macrocyclic Chemistry 2006, 56(1–2), 107–111.
7 Li, X.; Jin, Z.; Wang, J Complexation of allyl isothiocyanate by a- and b-cyclodextrin and its controlled release characteristics Food Chemistry 2007, 103(2), 461–466.
8 Vega-Lugo, A.C.; Lim, L.T Controlled release of allyl isothiocyanate using soy protein and poly(lactic acid) electrospun fibers Food Research International 2009, 42(8), 933–940.
9 Gunning, Y.M.; Gunning, P.A.; Kemsley, E.K.; Parker, R.; Ring, S.G.; Wilson, R.H.; Balake, A Factors affecting the release of flavor encapsulated in carbohydrate matrixes Journal of Agricultural and Food Chemistry 1999, 47(12), 5198–5205.
10 Yoshii, H.; Soottitantawat, A.; Liu, X.-D.; Atarashi, T.; Furuta, T.; Aishima, S.; Ohgawara, M.; Linko, P Flavor release from spray-dried maltodextrin=gum arabic or soy matrices as a function of storage
FIG 8 Chemical compensation between ln A and E of for the uncoated
complex powder () and the complex powders coated with PE1 (4), PE2
( 4
), and EVA ( ).
Trang 10relative humidity Innovative Food Science & Emerging Technologies
2001, 2(1), 55–61.
11 Shiga, H.; Yoshii, H.; Nishiyama, T.; Furuta, T.; Forssele, P.;
Poutanen, K.; Linko, P Flavor encapsulation and release
charac-teristics of spray-dried powder by the blended encapsulant of
cyclodextrin and gum arabic Drying Technology 2001, 19(7), 1385–
1396.
12 Soottitantawat, A.; Yoshii, H.; Furuta, T.; Ohgawara, M.; Forssell,
P.; Partanen, R.; Poutanen, K.; Linko, P Effect of water activity on
the release characteristics and oxidation stability of D -limonene
encap-sulated by spray drying Journal of Agricultural and Food Chemistry
2004, 52(5), 1269–1276.
13 Neoh, T.L.; Yoshii, H.; Furuta, T Encapsulation and release
charac-teristics of carbon dioxide in a-cyclodextrin Journal of Inclusion
Phenomena and Macrocyclic Chemistry 2006, 56(1–2), 125–133.
14 Dronen, D.M.; Reineccius, G.A Rapid analysis of volatile release
from powders using dynamic vapor sorption atmospheric pressure
chemical ionization mass spectrometry Journal of Food Science
2003, 68(7), 2158–2162.
15 Mortenson, M.; Reineccius, G.A Encapsulation and release of
men-thol, part 1: The influence of OSAn modification of carriers on the
encapsulation of L -menthol by spray drying Flavour and Fragrance
Journal 2008, 23(6), 392–397.
16 Mortenson, M.; Reineccius, G.A Encapsulation and release of
men-thol, part 2: Direct monitoring of L -menthol release from spray-dried
powders made with OSAn-substituted dextrins and gum acacia.
Flavour and Fragrance Journal 2008, 23(6), 407–415.
17 Mateus, M.L.; Lindinger, C.; Gumy, J.C.; Liardon, R Release
kinetics of volatile organic compounds from roasted and ground
cof-fee: Online measurements by PTR-MS and mathematical modeling.
Journal of Agricultural and Food Chemistry 2007, 55(25), 10117–
10128.
18 Bohn, D.M.; Cadwallader, K.R.; Schmidt, S Development and
vali-dation of dynamic vapor sorption–fast gas chromatography–flame
ionization detection method for rapid analysis of volatile release from
glassy matrices Journal of Agricultural and Food Chemistry 2005, 53(8), 3149–3155.
19 Bohn, D.M.; Cadwallader, K.R.; Schmidt, S Use of DSC, DVS-DSC, and DVS-fast GC-FID to evaluate the physicochemical changes that occur in artificial cherry Durarome 1
upon humidification Journal
of Food Science 2005, 70(2), E109–E116.
20 Normand, V.; Dardelle, G.; Bouquerand, P.E.; Nicolas, L.; Johnston, D.J Flavor encapsulation in yeasts: Limonene used as a model system for characterization of the release mechanism Journal of Agricultural and Food Chemistry 2005, 53(19), 7532–7543.
21 Zhao, Q.; Zhan, X.C.; Li, L.L.; Li, C.; Lin, T.; Yin, X.; He, N Pro-grammed humidifying in drug stability experiments Journal of Phar-maceutical Sciences 2005, 94(11), 2531–2540.
22 Li, L.L.; Zhan, X.C.; Tao, J.L Evaluation of the stability of aspirin in solid state by the programmed humidifying and non-isothermal experiments Archives of Pharmacal Research 2008, 31(3), 381–389.
23 Neoh, T.L.; Yamauchi, K.; Yoshii, H.; Furuta, T Kinetic study of thermally stimulated dissociation of inclusion complex of 1-methyl-cyclopropene with a-cyclodextrin by thermal analysis Journal of Physical Chemistry B 2008, 49(112), 15914–15920.
24 Neoh, T.L.; Yamamoto, C.; Ikefuji, S.; Furuta, T.; Yoshii, H Heat stability of allyl isothiocyanate and phenyl isothiocyanate complexed with randomly methylated b-cyclodextrin Food Chemistry 2012, 131(4), 1123–1131.
25 Paramita, V.; Iida, K.; Yoshii, H.; Furuta, T Effect of feed liquid tem-perature on the structural morphologies of D -limonene microencapsu-lated powder and its preservation Journal of Food Science 2010, 75(1), E39–E45.
26 Paramita, V.; Iida, K.; Yoshii, H.; Furuta, T Effect of additives on the morphology of spray-dried powder Drying Technology 2010, 28(3), 323–329.
27 Soottitantawat, A.; Yoshii, H.; Furuta, T.; Ohkawara, M.; Linko, P Microencapsulation by spray drying: Influence of emulsion size on the retention of volatile compounds Journal of Food Science 2003, 68(7), 2256–2262.