Modulated Release of a Volatile Compound from Starch Matrixes via

The release of a model volatile (diacetyl) from a system based on a starch ..... At low relative humidity the virtual absence of any release of diacet...
0 downloads 0 Views 130KB Size
Download PDF
Biomacromolecules 2002, 3, 305-311

305

Modulated Release of a Volatile Compound from Starch Matrixes via Enzymatically Controlled Degradation Gu¨lden Yılmaz,*,† Gaye O ¨ ngen,‡ Remy O. J. Jongboom,†,§ Herman Feil,† Cees van Dijk,† and Wim E. Hennink| ATO, Agrotechnological Research Institute, Bornsesteeg 59, P.O. Box 17, NL-6700 AA Wageningen, The Netherlands, Ege University, Faculty of Engineering, Department of Bioengineering, 35100 I˙ zmir, Turkey, and Department of Pharmaceutics, Utrecht Institute for Pharmaceutical Sciences (UIPS), Utrecht University, P.O. Box 80.082, 3508 TB Utrecht, The Netherlands Received August 14, 2001; Revised Manuscript Received November 20, 2001

The release of a model volatile (diacetyl) from a system based on a starch matrix, in which the volatile is dispersed, was studied. Kneading was used to obtain a homogeneous mixture (melt) composed of starch, glycerol R-amylase, and diacetyl. Samples were then ground to powders. When the starch powders were exposed to 30% relative humidity (RH) at 20 °C, no degradation of the starch matrix occurred. The samples only showed an initial burst release of diacetyl (around 10% of the loaded dose), whereas the remaining amount of diacetyl was not released, most likely due to the glassy character of the matrix and the low solubility of diacetyl in the matrix. However, when the samples were incubated at 90% RH, due to the uptake of moisture by the particles full release of the entrapped volatile occurred. The release of diacetyl from the matrix without enzyme followed first-order kinetics and, as expected, the release rate increased with decreasing particle size. Due to absorption of water, the enzyme became active and starch degradation occurred. The initial release of diacetyl from amylase-containing matrixes followed first-order kinetics as well. However, once the matrix was degraded to a certain extent, the particles collapsed, which was associated with concomitant rapid increase in release. The time at which the particle collapse occurred decreased with increasing enzyme concentration in the matrix. In conclusion, it is demonstrated that the release of a volatile from starch matrixes can be modulated both by the amount of coencapsulated matrix-degrading enzyme and by the humidity of the environment. Introduction Encapsulation technology represents an area of growing interest for controlled-release applications of, for example, flavors, fragrances, colorants, and bioactive compounds. Especially starch-based matrixes have provided promising and interesting properties in different application areas of encapsulation.1-4 Lately, methods involving thermomechanical processing techniques have been described for the preparation of starch matrixes. It has been shown that with these techniques the encapsulation of several ingredients in starch matrixes is possible, making cost-effective and safe products feasible.5-8 Both efficient and effective protection in combination with well-defined release characteristics of active ingredients are key requirements for the success of controlled-release systems.9,10 For matrix systems encapsulating a volatile compound, the release depends on several mutually dependent processes such as diffusion of the volatile compound through the matrix, transfer from the matrix to the environ* Corresponding author. E-mail address: [email protected]. † ATO, Agrotechnological Research Institute. ‡ Ege University. § Current address: Rodenburg Biopolymers, Denariusstraat 19, 4903 Rc Oosterhout, The Netherlands. | Utrecht University.

ment, and degradation/dissolution of the matrix material.9 The diffusion rate of a given compound through the matrix depends on the physicochemical properties of the matrix such as the presence of a crystalline phase, the Tg of the amorphous phase, and the presence of additives. Controlledrelease systems in which the release is partly governed by degradation have been described. Degradation of the matrix for these systems is, however, dependent on external parameters, such as environmental conditions such as pH and temperature as well as microbiological activity.11-14 Previously it has been shown that an enzyme coencapsulated in a delivery device can be used to degrade the matrix in an aqueous medium to result in a burst or delayed/pulsed release of an active agent.15-17 In contrast to the systems where degradation occurs upon incubation in an aqueous solution, we recently proposed a system in which the enzymatic degradation of the dry matrix with a matching enzyme was triggered after its exposure to a humidified atmosphere.18 This system is based on encapsulation of a thermostable R-amylase in a pregelatinized starch matrix and is obtained by means of kneading. It was observed that the R-amylase, a frequently applied enzyme in starch liquefaction,19-21 was successfully encapsulated in a starch matrix with almost full retention of its enzymatic activity. Degradation of the matrix was initiated by moisture uptake and occurred once the

10.1021/bm015600k CCC: $22.00 © 2002 American Chemical Society Published on Web 01/05/2002

306

Biomacromolecules, Vol. 3, No. 2, 2002

system was exposed to an atmosphere with a high relative humidity (90%). Further, the degradation rate of the starch in the matrix was controlled by the amount of encapsulated R-amylase. However, it has yet not been demonstrated that triggered degradation of this starch/amylase system can indeed be used as a control mechanism for the release of an encapsulated volatile. Thus, the objective of this study was to obtain insight into the factors affecting the release of a model volatile (diacetyl) dispersed in a starch matrix containing R-amylase as matrix degrading enzyme, with emphasis on the formulation and the environmental conditions. Materials and Methods Materials. Pregelatinized potato starch (amorphous, Flocgel LVW) containing 10% (w/w) water was purchased from AVEBE, The Netherlands. Glycerol was purchased from Chemproha Chemical Distributors BV, The Netherlands, and soy lecithin (Topcitin 50) was from Lucas Meyer, The Netherlands. The enzyme used in this study was a thermostable endoamylase (Termamyl 120L; produced by a genetically modified strain of Bacillus licheniformis, pH 5.6) supplied in a liquid formulation from Novo Nordisk A/S, Denmark. The activity of the liquid enzyme formulation (protein content was 38 mg/mL) was 120 KNU/g of enzyme formulation (1 Kilo Novo alpha-amylase Unit, KNU, activity unit is the amount of enzyme which breaks down 5.26 g of starch/h at 37 °C to dextrins and oligosaccharides at pH 5.6 (as given in the Termamyl product data sheet by Novo Nordisk). The model volatile, diacetyl (2,3-butanedione), was obtained from Merck (purity >98%). All other reagents used in the experiments were of analytical grade. Premix Preparation. Premixes of the matrix material (pregelatinized starch) and additives (glycerol served as plasticizer and lecithin served as lubricant and emulsifier) were prepared prior to processing. For preparing the premix, 44 g of pregelatinized starch, 8 g of glycerol and 1.2 g (3%) of lecithin were weighed and transferred into a glass beaker. The premixes were obtained by mixing these ingredients for 10 min at room temperature (25 °C) using a mixer (Hobart N-50, Ontario, Canada) equipped with a low shear spiralmixing tool at a speed of 70 rpm. Encapsulation Process and Sample Preparation. The encapsulation of both the enzyme and the volatile in different starch matrixes was performed using a kneading device (a lab scale torque rheometer, Haake-Rheocord 90, Haake Inc., USA, equipped with a Haake Rheomix 600 mixing device). The premixes (40 g) were introduced into the kneader and kneaded at a screw speed of 25 rpm and a barrel temperature of 40 °C until the premix was completely fed into the kneader. Next, the screw speed was increased to 80 rpm at a constant temperature of 40 °C. Ten minutes of kneading at 80 rpm was required to obtain a homogeneous melt of the matrix formulation. Subsequently, different amounts of the liquid enzyme formulation (0-1.2 mg of enzyme protein/g of dry starch, which corresponds with maximum of 3% w/w enzyme preparation; 70 ppm CaCl2 in glycerol was added to the enzyme formulation before being added

Yılmaz et al.

into the kneader) were added. Kneading was continued during the next 2 min, which resulted in a homogeneous starch melt as evidenced from the recovery of the torque to a constant value.22 Next, 1.6 mL of diacetyl (4% w/w, based on total composition) was added and kneading was continued for 3 min. When the processing was complete, the samples were removed from the kneader and cooled to room temperature. The samples were cryogenically ground in liquid nitrogen using a Polymix A-10 (Switzerland) grinder. Ground samples were sieved to obtain powders with three different particle sizes (d < 90 µm, 90 µm < d < 212 µm and 500 µm < d < 1000 µm), which were separately stored in closed glass containers at 4 °C before further analysis. Each formulation was prepared in duplicate starting from the premix preparation. Determination of the Dispersed Phase Morphology. The dispersed phase morphology of the samples was analyzed by scanning electron microscopy (SEM). Samples were fractured cryogenically and cross sections were analyzed as described previously.23,24 The SEM photos were digitized and processed for particle size analysis by using image-analyzing software (Image pro-plus). The data were converted into Excel 5.0 (Microsoft) in order to calculate their particle size distribution. Average diameters (n > 200, number mean D(1,0) and volume mean D(3,0)) of the dispersed diacetyl droplets and the polydispersity index (D(3,0)/D(1,0)) were calculated as described by Edmundson.25 Determination of Crystallinity in the Starch Matrixes. The presence of crystallinity in the starch matrixes was determined both directly after processing and after incubation of the samples for different times and relative humidity (RH) conditions. The samples which were incubated at 30% RH were analyzed directly, whereas the samples which were incubated at 90% RH were dried at 70 °C prior to analysis. In detail, the ground samples were placed in sample holders and were compacted. Next, the holders were placed in a Philips PC-APD diffractometer, model PW 3710, supplied with a Cu KR emitter (wavelength of 1.542 Å), operating at 50 mA and 40 kV. The samples were scanned in the angular range 5-40° (2θ) with a scan speed of 1.5 deg/min. Determination of R-Amylase Activity. To prepare the samples for enzyme activity measurement, 1 g of ground sample with a known amount of encapsulated R-amylase was added to 9 mL of 0.1 M sodium acetate buffer (pH 5.60). The resulting suspensions were continuously stirred at 4 °C until complete dissolution of the matrix and extraction of the enzyme was achieved (16 h). To clarify this solution, it was centrifuged using an Eppendorf centrifuge 5415 at 14 000 rpm for 3 min. The clear supernatant was used for the determination of the enzymatic activity using the iodine staining (amyloclastic) method.26 In detail, 0.5 mL of an adequately diluted supernatant sample was added to 5.0 mL of substrate solution (20 mg/mL of soluble starch in a 0.1 M sodium acetate buffer, pH 5.60). After 10 min of incubation at 25 °C, a sample of 0.5 mL was drawn and added to 5 mL of 0.1 M HCl to stop the enzymatic reaction. Next, 0.5 mL of this solution was added to an aqueous solution of KI (0.05 g of KI and 0.5 mg of I2 in 100 mL of demineralized water) and the intensity of the blue color was

Modulated Release of a Volatile Compound from Starch Matrixes

measured using a spectrophotometer (660 nm). The activity of amylase extracted from the different matrixes was compared with the activity of the enzyme as received from the supplier (120 KNU) and is given as percent of this initial activity. The activity measurements were performed after processing and during storage periods under different conditions at 20 °C and either 30% or 90% RH after 1, 2, 4, 8, 24, 48, and 72 h of incubation. Molecular Weight Determination of Starch. The average molecular weights and the molecular weight distributions of the starch originating from different samples (with and without encapsulated enzyme) were determined using a highperformance size-exclusion chromatography unit equipped with a multiangle laser-light-scattering detector and differential refractometer (HPSEC-MALLS-RI).27 In detail, 25 mg of ground sample and 2 mg of EDTA (to inhibit the enzyme activity) were dissolved in 2 mL of 0.25 M NaOH solution (10-15 min). This solution was then diluted in phosphate buffer (pH 8.0) to a starch concentration of 5 mg/ mL, filtered through a Millipore filter (0.45 µm), and injected into a sample loop (200 µL). Together with the eluant (phosphate buffer, 0.05 M pH 8.0), the sample was pumped through a guard column (TSK PWH-PE perd-guard, 7.5 × 7.5 mm, Beckman) followed by six main columns (Spherogel-TSK 1000 PW, 2000 PW, 3000 PW, 5000 PW, 5000 PWHR, and 6000 PW, 30 × 7, 5 mm, Beckman). Analysis of the eluant was performed with a Dawn-F muliangle laser photometer (Wyatt technology), equipped with an argonion laser operating at 488 nm using RI detectors at an angular range between 15 and 151° (Waters 410 differential refractometer). During the measurements the total system was kept at a temperature of 50 °C. A dn/dc value27 of 0.145 mL/g was applied to calculate the average molecular weights. Each sample was analyzed in triplicate. pH Measurements. The pH of the solutions obtained from the powdered samples as well as the samples that were incubated at different conditions (after time periods of 1, 2, 4, 8, 24, 48, and 72 h) was measured. The solutions were prepared by adding 1 g of a ground sample to 10 mL of demineralized water. The resulting suspensions were stirred for 2 h at room temperature. The pH of these solutions was measured using a Weilheim pH meter (pH 320, Germany). Encapsulation Efficiency and Release of the Model Volatile by Static Headspace Measurements. The encapsulation efficiency of diacetyl in the prepared starch matrixes and the release of this compound were determined using gas chromatography.28 A calibration curve was made by preparing dilutions of diacetyl in acetone ranging between 0.001 and 1 g/mL. This was achieved by transferring 1 mL of these dilutions into 40-mL headspace vials, which were immediately sealed with a crimp top aluminum cap and Tefloncoated septa. After the vials were loaded, they were incubated for 15 min at 40 °C and the diacetyl concentration in the headspace was measured. A headspace volume of 0.25 mL was injected into the GC in the split mode (1:100), using a HS850, CE Instruments auto sampler. Gas chromatography was performed using a HRGC MEGA2 series (FISON Instruments) equipped with a flame ionization detector and CP-WAX 52 CB semipolar column (50 m length, 0.32 mm

Biomacromolecules, Vol. 3, No. 2, 2002 307

i.d., 1.2 µm film thickness). The operation parameters of the chromatograph were as follows: injector temperature, 250 °C; detector temperature, 250 °C; column temperature (isothermal), 90 °C; carrier gas, He at 3 mL/min. The amount of sample used for the headspace analysis was selected such that the concentration of diacetyl in the gas phase was below saturation. The encapsulation efficiency of diacetyl was determined directly after processing using ground samples. The amount of diacetyl present in each sample was compared with a control, consisting of a mixture of 19 µL of diacetyl and 0.5 g of glycerol. In detail, 0.5 g of powdered sample was transferred into separate headspace vials, which were immediately sealed, incubated, and measured using GC as described above for the diacetyl dilutions in acetone. Encapsulation efficiency was given as the ratio of the initial peak area over the peak area of the control mixture of diacetyl and glycerol. To determine the diacetyl release, 5 g of a ground sample was placed in an aluminum Petri dish (d ) 5 cm) and left in open contact with the environment (a continuously ventilated room). The relative humidity was either 30% or 90% and the temperature was 20 °C. After storage periods of respectively 1, 2, 3, 4, 8, 24, 48, and 72 h, 0.5 g of the sample was transferred into a headspace vial which was immediately sealed, incubated, and analyzed for the amount of diacetyl. The percent release was calculated based on the amount of diacetyl, which was in the samples after processing and grinding. Each sample was analyzed in duplicate for encapsulation efficiency and release. Results and Discussion Preparation of Starch Matrixes Encapsulating R-Amylase and Diacetyl. Diacetyl and R-amylase were encapsulated in starch matrixes containing glycerol as plasticizer. First, a homogeneous melt consisting of starch, glycerol, and lecithin was obtained at a barrel temperature of 40 °C. The barrel temperature was kept constant throughout the process, and the registered melt temperature was between 40 and 45 °C. A melt could already be obtained at this relatively low temperature since the starch used in this study was pregelatinized and fully amorphous. Lecithin was used as an emulsifier and was not necessary for the formation of the melt. Although the amount of water was minimized, R-amylase started to break down the matrix during the encapsulation process. The degradation of the starch, as reflected by the decrease in average molecular weight, as expected18 increased with increasing amount of enzyme added (Figure 1). After the encapsulation process the activity of R-amylase was well preserved and amounted to 90 ( 5% of the initial activity. Both after storage and incubation at different conditions (30% and 90% RH), activity measurements of the encapsulated enzyme either in the absence or presence of diacetyl showed that the activity was almost quantitatively preserved (90 ( 5%). Therefore, it is concluded that diacetyl does not inactivate the enzyme. The starch particles were suspended in water before and after exposure to 90% RH

308

Biomacromolecules, Vol. 3, No. 2, 2002

Yılmaz et al.

Figure 1. Weight average molecular weight of starch as a function of the enzyme concentration after the kneading process (bars represent standard deviation (SD)).

for different times. The pH of the obtained solutions ranged between 5.5 and 5.7. Obviously, the buffering capacity of the enzyme preparation was adequate both to perform catalysis and to compensate for potential pH shifts caused for example by absorption of CO2 from the environment. The encapsulation efficiency of diacetyl was 80 ( 5% and was independent of the amount of enzyme. SEM analysis (data not presented) showed, in agreement with previous studies on related systems,23,24 the presence of a homogeneous matrix in which diacetyl was dispersed/embedded. The distribution of the dispersed phase of diacetyl in the matrix was comparable for all samples and the number average diameter of the encapsulated diacetyl droplets in the different matrixes ranged from 12 to 15 µm with a polydispersity between 1.13 and 1.25. X-ray analysis showed that starch in the matrixes was amorphous and the formation of a crystalline phase was not induced during processing. There was no indication of formation of inclusion complexes of neither lecithin nor diacetyl with amylose, which would have been revealed as a single helical crystalline phase.29 Moreover, lecithin due to its amphiphilc character is likely to be present at the oil/ starch interface. Indeed, in a previous publication we have demonstrated that the average size of the dispersed oil droplet was smaller in the presence of lecithin than in absence of this compound.24 Diacetyl forms a phase-separated system with starch/glycerol showing that the solubility of this compound in the continuous phase is low, indicating a low possibility of complex formation with starch. Degradation of Starch and Release of Diacetyl from Starch Matrixes at 20 °C and 30% RH. After exposure of the samples with a particle size 90 µm < d < 212 µm, containing different amounts of R-amylase, to 30% RH, 5-8% (w/w) of the water was lost and a new equilibrium was reached after about 8 h of incubation (data not shown). Gel permeation chromatography (GPC) analysis showed that under conditions of low relative humidity, the molecular weight of starch in the matrixes remained unchanged for at least a period of 14 days. This suggests that under these conditions the enzyme is not capable of performing any catalytic activity. Since the enzyme activity, measured after extraction of the enzyme from the matrixes, was preserved, the lack of catalytic activity is probably caused by the low

Figure 2. Release of diacetyl as a function of the enzyme concentration at 20 °C and 30% RH (particle size 212 < d < 500 µm, bars represent SD).

water activity within the matrix30 and the low molecular mobility of enzyme through the glassy starch matrix.31 From Figure 2 it can be seen that a burst release of diacetyl occurred during the first hours of incubation. In the absence of enzyme the release of diacetyl was 8% and increased with enzyme concentration up to 17% for the system with the highest R-amylase concentration (Figure 2). This burst release of diacetyl can most likely be ascribed to diacetyl droplets in contact with the surface as a result of grinding. At increasing enzyme concentrations more diacetyl is released as burst, probably due to enzyme-based degradation and weakening of the matrix, which has occurred during processing. After this initial burst of diacetyl release, no further significant release was measured. At low relative humidity the virtual absence of any release of diacetyl from the starch matrixes after the initial burst release may be explained by the low solubility of diacetyl in the matrix and low diffusivity through the glassy matrix. Degradation of Starch and Release of Diacetyl from Starch Matrixes at 20 °C and 90% RH. When the samples were stored at a relative humidity of 90%, the following observations were made. Due to this high relative humidity absorption of water occurred and the water content of the samples increased reaching a new equilibrium after about 10 h of incubation (Figure 3). In Figure 3 it can also be seen that the water uptake increased with increasing enzyme concentration. Furthermore it was observed that at increasing enzyme concentration the matrix further degraded as reflected by a decrease in the weight average molecular weight of starch (Figure 4). From this figure it can be seen that for all enzyme concentrations used, a fast decrease in molecular weight of starch occurred during the first 10 h, where after almost no further degradation occurred during the next 62 h. Besides the weight average molecular weight, also the molecular weight distribution curves (data not shown) did not change significantly after this first 10 h of incubation. At the end of this period (72 h), the molecular weight of the starch was lower at higher enzyme concentrations. Obviously the molecular weight of starch did not decrease to the same

Modulated Release of a Volatile Compound from Starch Matrixes

Biomacromolecules, Vol. 3, No. 2, 2002 309

Figure 5. Release of diacetyl from starch matrixes as a function of enzyme concentration (particle size 212 < d < 500 µm, 20 °C, RH ) 90%, bars represent SD).

Figure 3. Water uptake of starch matrixes with varying concentration of enzyme at 20 °C and 90% RH (particle size 212 < d < 500 µm, bars represent SD). The insert shows the moisture uptake during the first 8 h.

Figure 4. The weight average molecular weight of starch in time as function of enzyme concentration (particle size 212 < d < 500 µm; 20 °C, RH ) 90%, bars represent SD).

minimum value. During and after the incubation period of 72 h, full R-amylase activity was measured. Furthermore the pH of the system (pH 5.6) was preserved. This strongly suggests that the enzyme-dependent plateau values of the molecular weight of starch were not caused by a pH-induced inactivation of R-amylase. Three other possible explanations for the enzyme to stop performing catalytic activity can be given. First, due to water absorption of the matrix, starch crystallites are formed. The enzyme is then entrapped in amorphous regions surrounded by crystalline domains, resulting in a reduced mobility and thus catalytic activity of the enzyme. X-ray analysis, however, revealed that the matrixes initially were fully amorphous and, importantly, no crystallinity in the matrixes was detected in time once the matrixes were exposed to high RH. Since the samples were initially amorphous, the rate of recrystallization is obviously slow and does not occur within the time frame of the experiments. Second, upon exposure of the samples to 90% RH, the matrix absorbed water enabling the enzyme to

become catalytically active.30,32 As a consequence, the starch matrix starts being degraded and degradation products (low molecular weight starch fragments, i.e., oligosaccharides) accumulate in the matrix. The low molecular weight degradation products result in an enhanced water binding capacity33-35 (Figure 3), with a concomitant decrease in water activity. This low water activity can consequently cause the enzyme to be catalytically inactive. Third, besides that the formed oligosaccharides bind water, they might also act as an inhibitor of the enzyme. In conclusion, the arrest of enzyme activity after a certain incubation time of the matrix at 90% RH is caused either by a reduction in water activity or by product inhibition or by a combination of both factors. In contrast to the matrix incubated at 30% RH, the matrix incubated at 90% RH without enzyme showed a gradual release of the volatile (Figure 5). When the samples were incubated at 90% RH, water is rapidly absorbed (Figure 3), thereby changing the physical properties of the matrix in such a way that both the solubility of diacetyl in the matrix and the diffusivity due to the plasticizing effect of water cause release of the volatile from the matrix. In the absence of the matrix degrading enzyme the release of diacetyl from the matrix is proportional to the square root of time (first-order release kinetics,36 see Figure 5). In the presence R-amylase, the absorbed water acts as a trigger to start enzyme catalysis. At low enzyme concentrations (0.05 and 0.2 mg of protein/g of dry starch) the initial release of diacetyl was similar to the release from the matrix without enzyme. However, after about 1-2 h a sharp increase in the release of diacetyl was observed which could no longer be described by the first-order release kinetics. Obviously, the first-order release kinetics are overruled due to enzymeinduced changes in the physicochemical properties of the matrix. The water uptake and the final moisture content of these matrixes were dependent on the extent of starch degradation which was positively correlated with the amount of enzyme present in the matrix (Figures 3 and 4). Once the matrixes had absorbed a certain amount of water (between 10 and 20%), the matrixes softened and finally collapsed as a result of a combination of water uptake and the decreasing molecular weight of starch in the matrixes. Matrix collapse as a result of degradation and high water content can

310

Biomacromolecules, Vol. 3, No. 2, 2002

Figure 6. The relation between the initial (after-processing) molecular weight of starch in the matrix and both the percent diacetyl release during the first hour of incubation and water uptake (particle size 212 < d < 500 µm, 20 °C, RH ) 90%, bars represent SD).

Figure 7. Degradation of the starch matrix as a function of the particle size (0.05 mg/g of dry starch enzyme formulation, 20 °C, RH ) 90%, bars represent SD).

therefore very well explain the increased and rapid release of diacetyl. To support this hypothesis, a plot is shown relating both the percent diacetyl release and the water uptake during the first hour, against the initial molecular weight of starch (Figure 6). From Figure 6, it is obvious that an initial low release rate is observed at a starch molecular weight of >(6-7) × 105 g/mol. Below this molecular weight a steep increase in initial release rate is observed. Obviously for this system, the initial release rate is critical around this molecular weight of (6-7) × 105 g/mol and water uptake of >12%, under conditions of 90% RH. With reference to the initial water uptake, from Figure 6 it is obvious that the lower the molecular weight of the starch in the matrix, the faster and higher the water uptake. At decreasing molecular weights, the driving force of water uptake increases due to an increased osmotic value of the matrix. The Effect of Particle Size on Degradation of Starch and Release of Diacetyl at 20 °C and 90% RH. The effect of the particle size of the samples at a fixed enzyme concentration (0.05 mg of protein/g of dry starch) on both the degradation of starch and the release of diacetyl was studied. In Figure 7 the change in molecular weight of starch for different particle sizes as a function of incubation time is

Yılmaz et al.

Figure 8. Moisture uptake of starch matrixes of different particle sizes during 72 h of incubation at 20 °C, 90% RH (0.05 mg/g of dry starch enzyme formulation, bars represent SD).

Figure 9. The release of diacetyl release from starch matrixes as a function of the particle size (without enzyme, 20 °C, RH ) 90%, bars represent SD).

shown. It is observed that the smaller the particles, the faster the degradation. This is caused by a more rapid water uptake of smaller particles (Figure 8) due to their higher surfaceto-volume ratio which consequently resulted in a faster activation of the enzyme. From Figure 9 it can be concluded that without the inclusion of the enzyme, the release of diacetyl from smaller particles is faster and follows first-order kinetics. Once the enzyme is included, a sharp increase in release after the first 1-4 h of storage is observed. It can be seen that the sharp increase in release occurs after shorter incubation times when the particle size is smaller (Figure 10). This can be ascribed to a faster uptake of water by the smaller particles (Figure 8) which in turn is associated with faster activation of the enzyme. Conclusions In this study the concept of including a matrix-degrading enzyme in a formulation, consisting of starch with dispersed droplets of a volatile, was investigated to yield a delivery device characterized by first-order release kinetics followed by an increased release rate of the volatile due an enzymetriggered matrix degradation and collapse. The trigger for

Modulated Release of a Volatile Compound from Starch Matrixes

Figure 10. The release of diacetyl release of R-amylase containing starch matrixes (0.05 mg of enzyme/g of dry starch) as function of the particle size at 20 °C and RH ) 90% (bars represent SD).

the matrix degradation and collapse results from the absorption of water by the particles once exposed to a humidified atmosphere (90% RH). This water absorption causes the enzyme to become catalytically active. The release rate was strongly dependent on a combination of factors: the molecular weight of starch and the water content of the matrix, which both are strongly affected by the enzyme concentration. As a consequence, a very pronounced effect on the release of the volatile was observed as a function of the amount of enzyme in the matrix. This system opens promising possibilities for the design of delivery devices, which release their contents only once exposed to an atmosphere with a high relative humidity. References and Notes (1) Bergsma, J.; Wierik, G. H. P. In Starch 96sthe book; van Doren, H. A., van Swaaij, A. C., Eds.; Carbohydrate Research Foundation: Noordwijkerhout, 1997; pp 105-112. (2) Chen, J.; Jane, J. Effectiveness of granular cold-water-soluble starch as a controlled-release matrix. Cereal Chem. 1995, 72, 265-268. (3) Mauro, D. J. An update on starch. Cereal Foods World 1996, 41, 776-780. (4) Trimnell, D.; Shasha, B. S. Auto-encapsulation: A new method for entrapping pesticides within starch. J. Controlled Release 1988, 7, 25-31. (5) Carr, M. E.; Wing, R. E.; Doane, W. M. Encapsulation of Atrazine within a starch matrix by extrusion processing. Cereal Chem. 1991, 68, 262-266. (6) Doane, W. M. Encapsulation of pesticides in starch by extrusion. Ind. Crops Prod. 1993, 1, 83-87. (7) Trimnell, D.; Wing, R. E.; Carr, M. E.; Doane, W. M. Encapsulation of EPTC in starch by twin-screw extrusion. Starch/Sta¨ rke 1991, 43, 146-151. (8) Wing, R. E.; Carr, M. E.; Trimnell, D.; Doane, W. M. Comparison of steam injection cooking versus twin-screw extrusion of pearl cornstarch for encapsulation of chloroacetanilide herbicides. J. Controlled Release 1991, 16, 267-278. (9) Pothakamury, U. R.; Barbosa-Ca´novas, G. V. Fundamental aspects of controlled release in foods. Trends Food Sci. Technol. 1995, 6, 397-406. (10) Shahidi, F.; Han, X. Q. Encapsulation of food ingredients. Crit. ReV. Food Sci. Nutr. 1993, 33 501-547. (11) Arevalo-Niko, K.; Sandoval, C. F.; Galan, L. J.; Imam, S. H.; Gordon, S. H.; R. V. Greene, Starch-based extruded plastic films and evaluation of their biodegradable properties. Biodegradation 1996, 7, 231-237. (12) Dave, H.; Rao, P. V. C.; Desai, J. D. Biodegradation of starchpolyethylene films in soil and by microbial cultures. World J. Microbiol. Biotechnol. 1997, 13, 655-658.

Biomacromolecules, Vol. 3, No. 2, 2002 311 (13) Imam, S. H.; Chen, L.; Gordon, S. H.; Shogren, R. L.; Weisleder, D.; Greene, R. V. Biodegradation of injection molded starch-poly (3-hydroxybutyrate-co-3-hydroxyvalerate) blends in a natural compost environment. J. EnViron. Polym. Degrad. 1998, 6, 91-98. (14) Whitney, P. J.; Swaffield, C. H.; Graffham, A. J. The environmental degradation of thin plastic films. Int. Biodeterior. Biodegrad. 1993, 31, 179-198. (15) Dumolin, Y.; Cartilier, L. H.; Mateescu, M. A. Cross-linked amylose tablets containing R-amylase: an enzymatically controlled drug release system. J. Controlled Release 1999, 60, 161-167. (16) Fishel-Ghodsian, F.; Brown, L.; Mathiowitz, E.; Brandenburg, D.; Langer, R. Enzymatically controlled drug delivery. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 2403-2406. (17) Hennink, W. E.; Talsma, H.; Borchert, J. C.; De Smedt, S. C.; Demeester, J. Controlled release of proteins from dextran hydrogels. J. Controlled Release 1996, 39, 47-55. (18) O ¨ ngen, G.; Yılmaz, G.; Jongboom, R. O. J.; Feil, H. Encapsulation of R-amylase in a starch matrix. Carbohydr. Polym., in press. (19) Likimani, T. A.; Sofos, J. N.; Maga, J. A.; Harper, J. M. Extrusion cooking of Corn/Soybean Mix in the Presence of Thermosbable R-Amylase. J. Food Sci. 1991, 56, 99-105. (20) Hakulin, S.; Linko, Y.-Y.; Linko, P.; Enzymatic Conversion of Starch in Twin-screw HTST-Extruder. Starch/Sta¨ rke 1983, 35, 411-414. (21) Toma´s, R. L.; Oliveira, J. C.; McCarthy, K. L. Influence of Operating Conditions on the Extend of Enzymatic conversion of Rice Starch in Wet Extrusion. Lebensm.-Wiss.-Technol. 1997, 30, 50-55. (22) Wilson, A. J.; Wooding, A. R.; Morgenstern, M. P. Comparison of work input requirement on laboratory-scale and industrial-scale mechanical dough development mixers. Cereal Chem. 1997, 74, 715-721. (23) Yılmaz, G.; Jongboom, R. O. J.; Van Soest, J. J. G.; Feil, H. Effect of glycerol on the morphology of starch-sunflower oil composites. Carbohydr. Polym. 1999, 38, 33-39. (24) Yılmaz, G.; Jongboom, R. O. J.; Feil, H.; Hennink, W. E. Encapsulation of sunflower oil in starch matrices via extrusion: effect of the interfacial properties and processing conditions on the formation of dispersed phase morphologies. Carbohydr. Polym. 2001, 45, 403410. (25) Edmundson, I. C. Particle-Size Analysis. In AdVances in Pharmaceutical sciences; Bean, H. S., Beckett, A. H., Canless, J. F., Eds.; Academic Press: London 1967; Vol. 2, pp 95-179. (26) Robyt. J. F. Enzymes in the hydrolysis and synthesis of starch. In Starch: Chemistry and Technology, 2nd ed.; Whistler, R. L., Bemiller, J. N., Paschall, E. F., Eds.; Academic Press: Orlando, San Diego, New York, Austin, Boston, London, Sydney, Tokyo, Toronto, 1984; pp 87-123. (27) Veelaert, S. Dialdehyde starch: Production, properties and applications. Ph.D. Thesis, Faculteit Landbouwkundige en Toegepaste Biologische Wetenschappen 1996, The Netherlands. (28) Techniques for analyzing food aroma; Marsili, R., Ed.; Marcel Dekker Inc.: New York, Basel, Hong Kong, 1997; pp 27-59. (29) van Soest, J. J.; Hulleman, S. H. D.; de Wit, D.; Vliegenthart, J. F. G. Crystallinity in starch bioplastics. Ind. Crops Prod. 1996, 5, 1122. (30) Reed, G. In Food Science and Technology; Anson, M. L., Chichester, C. O., Mrak, E. M., Stewart, G. F., Eds.; Academic Press: San Diego, New York, Boston, London, Sydney, Tokyo, Toronto, 1966; pp 2930. (31) Lourdin, D.; Coignard, L.; Bizot, H.; Colonna, P. Influence of equilibrium relative humidity and plasticizr concentration on water content and glass transition of starch materials. Polymer 1997, 38, 5401-5406. (32) Enzymes, Eagan Press Handbook Series; Mathewson, P. R., Ed.; Eagan Press: St. Paul, MN, 1998; pp 1-13. (33) Qi, Z. H.; Xu, A. Starch based ingredients for flavor encapsulation. Cereal Foods World 1999, 44, 460-465. (34) Pessa, E.; Suortti, T.; Autio, K.; Poutanen, K. Molecular weight characterization and gelling properties of acid-modified maize starches. Starch/Sta¨ erke 1992, 44, 64-69. (35) Ratkovic, S.; Pissis, P. Water Binding to Biopolymers in Different Cereals and Legumes: Proton NMR Relaxation, Dielectric and Water Inhibition Studies. J. Mater. Sci. 1997, 32, 3061-3068. (36) Controlled release of biologically actiVe agents; Baker, R., Ed.; John Wiley & Sons: New York, Chichester, Brisbane, Toronto, Singapore, 1987; pp 39-83.

BM015600K