Journal of Colloid and Interface Science 450 (2015) 182–188

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier .com/locate / jc is

Barrier properties of heat treated starch Pickering emulsions

http://dx.doi.org/10.1016/j.jcis.2015.03.0040021-9797/� 2015 Elsevier Inc. All rights reserved.

⇑ Corresponding author at: Department of Food Technology, Engineering andNutrition, Lund University, Box 124, SE 221 00 Lund, Sweden. Tel.: +46 46 222 8311;fax: +46 46 222 4622.

E-mail addresses: Malin.Sjoo@food.lth.se (M. Sjöö), sinan_cem.emek@biochemistry.lu.se (S.C. Emek), hall_tina_87@hotmail.com (T. Hall), Marilyn.Rayner@food.lth.se(M. Rayner), Marie.Wahlgren@food.lth.se (M. Wahlgren).

Malin Sjöö a,b,⇑, Sinan Cem Emek c,d, Tina Hall a, Marilyn Rayner a, Marie Wahlgren a

a Department of Food Technology, Engineering and Nutrition, Lund University, Box 124, SE 221 00 Lund, Swedenb Speximo AB, Medicon Village, SE 221 83 Lund, Swedenc Department of Biochemistry and Structural Biology, Center for Molecular Protein Science, Lund University, Box 124, SE 221 00 Lund, Swedend Greenleaf Medical AB, Medicon Village, SE 221 83 Lund, Sweden

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 20 October 2014Accepted 4 March 2015Available online 11 March 2015

Keywords:Barrier propertiesEncapsulationGelatinizationLipolysisPickering emulsionStarch granule

a b s t r a c t

Hypothesis: There is a recognized technological need for delivery systems encapsulating lipophilic sub-stances in food and pharmaceutical products. Pickering emulsions can provide well-defined and highlystable systems, but may not provide good enough barrier properties. Starch granules, recently being usedfor Pickering stabilization, have the advantage of the ability to swell during gelatinization. Hence, thisproperty could be used to tune and control barrier properties.Experiments: Oil-in-water Pickering emulsions stabilized by starch were subject to heat treatment at dif-ferent conditions. The influence of temperature, time, and storage on emulsion drop characteristics wasevaluated. In order to further evaluate the barrier properties, lipolysis using the pH-stat method wasapplied and the effect of starch concentration, treatment temperature, and preliminary oral conditionswere also investigated.Findings: A better encapsulating barrier was obtained by starch swelling at the oil drop interface. Thiswas seen as reduced lipase activity. The internal oil drop size remained intact and the starch was keptat the interface during heat treatment. The extent of swelling could be controlled by the heating condi-tions and had impact on the ability to prevent lipase transport through the starch barrier layer. Additionof a-amylase simulating oral digestion only had minor impact on the barrier effect.

� 2015 Elsevier Inc. All rights reserved.

M. Sjöö et al. / Journal of Colloid and Interface Science 450 (2015) 182–188 183

1. Introduction

There is a recognized technological need for oral delivery sys-tems that encapsulates, protects and gives controlled release ofbioactive substances in food and pharmaceutical products [1].Emulsions [1] or dried emulsions [2], either oil/water for lipophilicsubstances or double emulsions (w/o/w) for hydrophilic sub-stances, can be interesting candidate formulations in this respect.However, such systems have a large surface area and will be sus-ceptible to surface induced degradation, such as lipid digestion.However, by structuring the interface between oil and water onecould achieve delayed lipid digestion and for example induce sati-ety [3,4], or targeted and controlled release of bioactive compo-nents within the gastro-intestinal tract [1,5]. Pickering emulsionshave been seen to at least in some cases reduce lipid digestion inemulsions [6].

Emulsions made without the use of traditional emulsifiers hasbeen of interest in application areas where the use of surfactantcauses adverse effects such as air entrapment, foaming, irritancy,or biological interactions [7]. An elegant way to avoid surfactantuse is to substitute emulsifiers by solid particles as drop stabilizingagents. Due to their distinctive characteristics and promisingtechnological applications [8], the use of particles to stabilizeemulsions has attracted substantial research interest. Particle sta-bilized emulsions, so called Pickering emulsions first observed over100 years ago [9,10], are known to possess a high degree of stabil-ity even without the addition of surfactant [11,12]. A more detaileddescription of Pickering emulsions can be found in some recentreviews [13,14].

Lately starch granules have been studied for of use in Pickeringemulsions [15–19]. Starch granules are biodegradable and wellaccepted for use in both food and pharmaceutical applications,even after hydrophobic modification with <3% octenyl succinicanhydride (OSA) [20]. Hydrophobic modification of starch isrequired in order to obtain affinity for the lipid phase. Starch natu-rally varies in size from 0.5 to almost 100 lm depending on thebotanical source [21]. Previously, intact starch granules from qui-noa (Chenopodium quinoa Willd.) was shown to efficiently stabilizeoil-in-water emulsions [13,19]. This ability was attributed to thesmall granule size (0.5–2 lm), unimodal size distributions and dis-tinctive morphology, i.e. rounded polyhedrons.

Due to the relatively large size of the stabilizing starch granules,at least in comparison to emulsifiers like surfactant molecules(0.4–1 nm), and globular proteins (1–5 nm) [8], there are relativelylarge spaces between them even if assuming a fully covered inter-face. Thus even though the starch granules are stabilizing theemulsion drop interface against coalescence, they are not likelyproviding equally good barrier properties. A huge advantage ofusing intact starch granules over other particles is therefore starchphysicochemical properties in terms of gelatinization. This irrever-sible process, as defined by Atwell et al. [22] includes granularswelling, native crystalline melting, loss of birefringence, and sol-ubilization, leading to collapse of molecular order. Depending onsubsequent storage conditions, the gelatinized starch may thenbe further structured by recrystallization. Heat treatment causingstarch gelatinization in Pickering emulsions has previously beenshown to just slightly increase the emulsion drop size distributionalthough also introducing drop aggregation [2].

One method to quantify the relative impenetrability of an inter-facial barrier for comparing differently treated emulsions is tomeasure the ability to decrease the rate of lipolysis [23]. The diges-tion of lipids is an interfacial process that involves the interactionof the lipase enzyme and its co-factors with the surface of thedrops such that the enzyme can come into close contact with itssubstrate [1]. For this reason the interfacial area, i.e. the specificsurface area of the emulsion is of importance and is given by:

S ¼ 6ud32

where S is the surface area per unit volume of emulsion (m2), u isthe oil volume fraction, and d32 is the Sauter mean diameter. Therate of digestion is often monitored using the pH-stat method[1,5,23]. This method partially mimic the physiological conditionsin the small intestine although amylases known to break downstarch are not generally included.

The aim of the present study was to modify the interfacial bar-rier properties of starch stabilized Pickering emulsion drops bypartially or completely gelatinizing starch at the interface. Theinfluence of starch surface packing, i.e. starch concentration, afterheating to a specific temperature was evaluated. Furthermore,the influence of different degrees of gelatinization was studied asobtained by heating emulsions of a specific composition to differ-ent temperatures for selected periods of time. The focus was toevaluate the ability of the starch barrier to protect the lipid phasein general and not specifically to mimic gastrointestinal (GI) condi-tions. To quantify relative changes in initial rate of lipolysis as ameasure of barrier properties the pH stat lipolysis method wasused for analysis. This method was selected since the aim wasnot to quantify absolute bioavailability of the oil phase. However,to further mimic conditions at oral digestion, a-amylase was addedin some experiments.

2. Experimental

2.1. Materials

The starch granules were isolated from quinoa grains (Biofood,Sweden) and chemically modified by OSA (2.9%) as described pre-viously [24]. OSA donated by Lyckeby Culinar AB, Sweden was usedfor hydrophobic modification. Two different oils, paraffin (MerckChemicals, Germany, 1.07174) and MCT-oil, Miglyol 812 (Sasol,Germany, 050 223), were used. All other chemicals was of PA qual-ity and obtained from Sigma–Aldrich. The continuous phase ofemulsions was a 5 mM phosphate buffer with pH 7 containing0.2 M NaCl. A 1 mM Tris maleat buffer (pH 7.0) with 4 mMNaTDC, 1 mM CaCl2 and 0.15 M NaCl was used as assay buffer forthe lipolysis. Porcine pancreas lipase and co-lipase (L0382-100KUand C3028) were dissolved, 1 mg/ml, in assay buffer and aqueoussolution, respectively. In some experiments a-amylase (A6814)was used at the concentration 100 U/ml.

2.2. Preparation of emulsions

Emulsions were prepared in glass test tubes by mixing with anYstral (D-79282, Ballrechten-Dottingen, Germany) at 22000 rpmfor 30 s. For studying the influence of starch concentration, emul-sions with 2.7 ml of the continuous phase, 0.3 ml of the oil phase(10%) and 22.5–180 mg starch (75–600 mg starch per ml oil) wereprepared. All other emulsions were prepared using 7% oil phaseand 214 mg starch per ml oil.

After emulsification, emulsions were heat treated in a waterbath at 45–100 �C. In order to compare different degrees ofgelatinization the holding time at the desired temperatures wasvaried from 1 to 30 min. As comparison, starch was also heat trea-ted in buffer and cooled to room temperature (22 �C) before emul-sification. For lipolysis experiments the holding time was 1 min,and for experiments evaluating the influence of starch concentra-tion, emulsions were heated to 70 �C.

In one storage experiment, heat treated emulsions with paraffinoil as dispersed phase were stored at 5 �C for 16 h and then kept atroom temperature for 24 h, or as comparison kept only at roomtemperature for 48 h and 8 weeks, respectively. Emulsions used

30 50 70 90 110 130

← E

ndot

herm

al h

eat f

low

, mW

Temperature, °C

Fig. 1. DSC thermogram for gelatinization of quinoa starch in buffer. Average valuesand standard deviations (n = 3) of gelatinization peak temperature (Tp) andtransition enthalpy (DH) included for starch:buffer (1:3), starch:buffer:MTC oil(1:3:1), and starch:buffer:paraffin oil (1:3:1), respectively.

184 M. Sjöö et al. / Journal of Colloid and Interface Science 450 (2015) 182–188

for lipolysis experiments were generally analyzed after overnightstorage, some samples were also analyzed after 1 and 4 weeks.

2.3. Differential scanning calorimetry (DSC)

Quinoa starch gelatinization properties were analyzed using aSeiko 6200 DSC (Seiko Instruments Inc. Shizuoka, Japan) calibratedwith indium (Mp = 156.6 �C). Starch and buffer was weighed intocoated aluminum pans (TA Instruments, New Castle, USA) at a ratioof 1:3 in order to determine the gelatinization transition enthalpy(DH, J/g dry matter) and peak temperature (Tp, �C). Additionally,starch:buffer:oil mixtures were analyzed at the ratio 1:3:1. Anempty pan was used as reference. The scanning rate was10 �C/min from 6 to 140 �C.

2.4. General emulsion characterization

The particle size distributions of starch particles and emulsiondrops were measured using laser diffraction in a flow through cellusing Coulter LS 130 (Coulter Electronics Ltd, Luton Beds, England)for emulsions varying in starch concentration, and Mastersizer2000 (Ver.5.60, Malvern, United Kingdom) for all other samples.A small volume of sample was added to the flow system andpumped through the optical chamber for measurements. Therefractive index of the particle was set to 1.54 (corresponding tothe starch covering the drops), and the refractive index of the con-tinuous phase was set to 1.33 (RI of water).

Emulsions were further observed under a light microscope(Olympus BX50, Tokyo, Japan) after 5 times dilution with bufferand without the use of cover glass. In order to follow starchgelatinization in terms of loss of birefringence a polarization filter(U-ANT, Olympus) and a wavelength filter (U-TP530, Olympus,Tokyo, Japan) was used. A 1280 � 960 resolution camera (DFK41AF02, The Imagingsource, Germany) was connected to themicroscope and images were processed using the Java image pro-cessing program ImageJ (version 1.42 m). Pictures were analyzedin order to determine the gelatinized starch layer and the diameterof the oil drops, respectively, 20 drops of each sample were mea-sured and compared with laser diffraction data. The average oildrop size based on laser diffraction data was calculated as the peakvalue of emulsion drops with twice the mean starch particle sizewithdrawn. This correspond to withdrawal of the average barrierlayer of a cross-section, assuming a monolayer of starch aroundeach drop.

2.5. pH-stat method

The pH-stat method was performed with slight adjustmentsaccording to Albertsson et al. [25]. The activity of lipase and coli-pase was determined by pH-stat titration using a TIM854 modelRadiometer (Analytical SAS, Cedex, France). The sample, emulsionor control, was stirred with 15 ml assay buffer and 3 ll of lipaseand colipase, respectively. The pH was maintained at 7.0 by titra-tion of 0.1 M NaOH and the consumption rate (ml/s) after 30 minwas used to determine the activity of lipase and colipase.

In this study, the main aim was to investigate the relativechanges in initial rate of lipolysis as a measure of barrier propertiesrather than absolute bioavailability of the oil phase. Thereby, thelipase activity of the most easily accessible oil interface, i.e. thatof unheated emulsions was set as 100% and used as control forevery set of lipolysis experiments. Then the relative decrease inactivity in heat-treated starch granule stabilized emulsions wasdetermined. As additional control, samples containing only bufferor buffer and starch were analyzed for lipase activity in theabsence of oil. In order to perform preliminary experiments at con-ditions slightly more relevant to oral intake, a-amylase was added

to emulsion which was then stirred for 2 min at 37 �C before thelipolysis experiment. This would roughly correspond to the timein mouth.

3. Results and discussion

3.1. Starch gelatinization properties and effects on emulsion drops

When heated in the presence of water, starch granules swelland gelatinize. A quinoa starch gelatinization endotherm as deter-mined by DSC in the excess of water is seen in Fig. 1. Neither thegelatinization temperature nor the transition enthalpy wereaffected by the presence of oil since no differences were foundfor starch either in the presence of buffer only, or in emulsionswith buffer and paraffin or MCT oil, respectively, see Fig. 1. Thebarrier layer formation caused by gelatinization at the interfaceof heat treated starch stabilized emulsions was also independentof the oil phase used. Starch stabilized emulsion drops increasedin size with increased temperature treatment (Fig. 2). This increaseupon heating was directly caused by the swelling of the starchlayer whereas the internal oil drop size remained intact. As seenin Fig. 2 the starch layer thickness measured from micrographscorresponded to a calculated starch layer thickness based on emul-sion drop size with the average internal oil drop size withdrawn.The internal oil drop size, d32 = 25.1 lm, was calculated, based onlaser diffraction measurements, as the emulsion drop size inunheated emulsions with the intact starch granule barrier sub-tracted. Since the increased emulsion drop size upon heating, asmeasured by light scattering and also seen under microscope,was caused by starch swelling, it was concluded that the heattreatment did not cause coalescence. The emulsion stability wasfurther demonstrated by the absence of free oil appearing on topof samples upon heating or subsequent storage.

Heating time also affects starch swelling. The influence ofgelatinization temperature and time on starch swelling was fol-lowed by the increase in drop size (Fig. 3). It was clear that thestarch swelling increased with time mainly for temperaturescorresponding to the melting interval in the DSC thermogram(Fig. 1), whereas at temperatures above this the rate of swellingwas much higher, causing a maximum increase of the starch layerat the drop interface already after 1 min heating to 90 �C or more.Thereby the temperature mainly affected the emulsion drop sizesat short treatment times, assuming that the temperature was high

0

5

10

15

20

25

30

35

40

45

0 20 40 60 80 100 120

Parti

cle

size

(d32

µm)

Single starch layer

Starch stabilized emulsions

Temperature (°C)

Fig. 2. Heat treated samples (1 min) as a function of temperature. Quinoa starch stabilized paraffin oil emulsion drops (diamonds) measured by light scattering (averages ofmode d32 with standard deviation, n = 3), the single starch layer thickness measured from micrographs with standard deviation, n = 20 (squares), and calculated from theemulsion drop size using light scattering data (filled diamonds), respectively. The bars in the inserted micrographs highlights the starch layer thickness at differenttemperatures.

0

10

20

30

40

50

60

70

0 10 20 30 40

Dro

p si

ze (d

43µm

)

Heat treatment time (min)

Fig. 3. Emulsion drop size distribution as a function of heat treatment time (1, 5, 10,30 min) for heating to 45 (diamonds), 55 (squares), 65 (triangles), 75 (circles), 90(�), and 100 (+) �C, respectively.

M. Sjöö et al. / Journal of Colloid and Interface Science 450 (2015) 182–188 185

enough to efficiently gelatinize starch (>45 �C). This resulted in theselection of 1 min treatment for the samples used for lipolysisexperiments.

Heating of emulsions was further seen to induce only partialgelatinization of starch granules at the drop interface. The starchgelatinization was expected to be initiated, and throughout theprocess be more pronounced, at parts of granules facing the

Fig. 4. Micrographs of unheated (a and b) and heated (70 �C) (c and d) emulsions showingdrop surface. Light microscopy (a and c) and birefringence seen under polarized light (b

continuous aqueous phase. Although the boundary between theindividual granules thereby became diffuse, there was still a cer-tain degree of ordered structure within the starch at the oil inter-face as seen by residual birefringence, Fig. 4. The birefringence ofstarch granules was very clear in unheated emulsions where thestarch granules are still crystalline, whereas in heated samples,only birefringence of starch granule parts situated at the oil inter-face was visible under polarized light. Birefringence remainedthroughout the heating process, even at 100 �C. Although birefrin-gence does not prove a crystalline structure per se, it is an indica-tion of an oriented structure within the granule remnants.Thereby the particle stabilization mechanism could be remainedat the interface [2] while at the same time achieving a more com-pact outer layer towards the aqueous phase, which could provideenhanced barrier properties for heat-treated starch stabilizedemulsions.

In a separate experiment starch was pre gelatinized by heatingto corresponding temperatures before emulsification as compari-son. When pre gelatinized starch was used for emulsification, lar-ger emulsion drops were obtained with starch heated to 55 �C for1 min, whereas when heated to higher temperatures the emulsiondrop size was decreased (Table 1). When heated to 45 �C, no swel-ling has occurred and the emulsions were identical. At 55 �C theeffective particle size of the swollen granules was increased. Itwas also previously observed that when using the same starch con-centrations, larger particles cause larger emulsion drops to beformed [26]. The decrease in the measured emulsion drop sizeupon heating starch to higher temperatures prior to emulsificationcan be indicative of emulsions stabilized not by intact granules but

swelling of the gelatinized starch layer and remaining starch birefringence at the oiland d).

Table 1Drop size as mode of d43 (lm) for emulsions heat treated to different temperatures and stored at different conditions, and for emulsions prepared with starch pre-heated prior toemulsification. Mode d43 is the peak of the emulsion drop size from the volume based drop size distribution. Averages and standard deviations, n = 3.

T (�C) Starch gelatinized in emulsion Pre-gelatinized

Not stored 16 h/5 �C 16 h/5 �C + 24 h/22 �C 48 h/22 �C 8 weeks/22 �C Not storedMode d43 (lm) Mode d43 (lm) Mode d43 (lm) Mode d43 (lm) Mode d43 (lm) Mode d43 (lm)

40 36.7 ± 1.645 34.8 ± 0.0 37.6 ± 0.0 37.6 ± 0.0 40.5 ± 0.0 34.8 ± 0.0 43.7 ± 0.050 39.7 ± 4.555 43.8 ± 3.3 44.9 ± 2.0 44.9 ± 2.0 41.6 ± 1.8 36.9 ± 6.0 66.1 ± 10.160 42.8 ± 3.865 46.0 ± 2.0 43.8 ± 3.3 43.8 ± 3.3 44.9 ± 2.0 37.0 ± 12.3 30.0 ± 2.370 48.4 ± 2.275 47.3 ± 3.6 52.3 ± 2.3 52.2 ± 2.3 49.7 ± 2.2 39.6 ± 10.4 27.0 ± 1.280 49.7 ± 2.290 52.8 ± 6.5 54.0 ± 8.8 56.6 ± 6.7

100 55.1 ± 5.9 54.9 ± 0.0 56.4 ± 2.5

186 M. Sjöö et al. / Journal of Colloid and Interface Science 450 (2015) 182–188

by starch molecules leaching from the granule or by presence offragments of gelatinized starch disrupted during the emulsificationprocess. Thus, the mechanism for emulsion stability obtainedwhen heating the starch when already present at the drop inter-face clearly differed from starch gelatinized before emulsification.

Heat treated emulsions were further subject to cooling andeventually further re tempering to room temperature, as well asto prolonged storage at room temperature, Table 1. Such tempera-ture cycling has previously been found to facilitate amylopectinrecrystallization [27,28]. Amylopectin recrystallization mayincrease the stability of the starch barrier layer towards enzymaticdigestion [29]. Although the temperature cycling treatment wasexpected to affect the structure of the starch barrier at a molecularlevel, it did not affect the macroscopic properties as determined byparticle size distribution (Table 1). The emulsion was also stableduring further storage as seen in Table 1. After 8 weeks storage,the emulsion oil drop size was seen intact under microscope. Forthe total drop size, however, a slight reduction was detected bylaser diffraction for heated emulsions (Table 1). This indicateschanges in the gelatinized starch barrier that could be due todegradation or rearrangement of the starch molecules at the inter-face. However, no preservatives were added to the system thusthere is a clear risk of bacterial amylase being present in thesamples.

0

20

40

60

80

100

120

30

32

34

36

38

40

42

44

46

48

50

0 20 40 60 80

Lipa

se a

ctiv

ity (%

)

Dro

p si

ze (d

43µm

)

Temperature (°C)

Fig. 5. Lipase activity (filled circles) and emulsion drop size before (diamonds) andafter (gray diamonds) lipolysis as a function heat treatment temperature ofemulsion at the concentration 214 mg starch per ml oil.

3.2. Effect of starch swelling on barrier properties as measured bylipolysis

The starch swelling obtained by temperature treatment clearlyaffected the barrier properties (Fig. 5). When starch was partiallygelatinized, the layer around the emulsion drops became moreimpermeable towards lipase activity as compared to unheatedemulsions. Heating to 55 �C and above caused enough starch swel-ling to better cover the oil interface. Non swelled particles would,even at close packing, have space between the individual particles,making the oil interface easily accessible for the lipase. Theincreased stability towards lipase upon heating was further con-firmed by the drop size after lipolysis that was more intact com-pared to the original emulsion when heated to the highertemperatures (Fig. 5). The oil phase degradation generally causeddecreased emulsion drop size after lipolysis. The increased emul-sion drop size upon heating measured before lipolysis (Fig. 5)was caused by starch swelling at the drop interface as previouslydescribed. The peak value of the drop size distribution was gener-ally used since the average drop size of heated emulsions showed afurther increase that could be referred to drop aggregation.Microscopic images clearly showed that the gelatinized starchlayer was still attached to the drop surface. Moreover, the partialremaining starch birefringence at the oil drop surface as previouslyshown in Fig. 4 was a strong indication of remaining particle char-acter at the drop interface. The emulsions shown in Fig. 5 wereanalyzed by the pH-stat method also after 4 weeks of storageand the protective effect of the gelatinized layer was found intact.

With increased starch concentration, the barrier layer at theinterface became denser as a result of closer packing of starchgranules. This was seen as reduced lipolysis rate with increasedstarch concentration (Fig. 6) when heated to 70 �C, which was atthe end of the gelatinization endotherm obtained by DSC (Fig. 1).The emulsion drop size decreased with increased starch concentra-tion (Fig. 6) which was expected based on previous findings [18].As a result of the reduced drop size, the total area of the emulsionincreased with starch concentration, thereby increasing the areaavailable for enzymatic attack. Therefore, the lipase activity wasscaled by the specific surface area in order to compare the barrierproperties of the starch layer in the emulsions as has also beenreported previously [19]. At the highest starch concentration(600 mg/ml oil), the lipase activity was reduced by 65.6%, and evenat 150 mg/ml the reduction was 57.4%. This was interpreted ashigh capacity of the starch to cover the oil drop surface. Even atthe lowest starch concentration, 75 mg/ml oil, when there wasnot enough starch available to completely cover the surface, thelipase activity was reduced by 31.9%. The large difference betweenthe two lowest concentrations could therefore be interpreted as

0

10

20

30

40

50

60

0

10

20

30

40

50

60

70

80

90

100

0 200 400 600 800

Dro

p si

ze (d

32µm

)

Rel

ativ

e lip

olys

is ra

te (%

)

Starch concentration (mg/ml oil)

Fig. 6. Relative lipolysis rate (filled circles) and emulsion drop size (diamonds) as afunction of starch concentration in emulsions heated to 70 �C for 1 min, averagesand standard deviations, n = 2.

M. Sjöö et al. / Journal of Colloid and Interface Science 450 (2015) 182–188 187

the ability of the swollen starch to better cover the oil dropinterface.

3.3. Effect of addition of a-amylase

Amylase is known to digest starch, and gelatinized starch isgenerally more susceptible to digestion. Therefore, an experimentwith a-amylase addition to the emulsion for 2 min before lipoly-sis was performed. This would roughly correspond to an oraltreatment. The a-amylase activity in the human saliva has beenmeasured to vary between 10 and 160 IU/ml depending on themethod of collection, the saliva flow and the satiety of the person[30,31]. In this experiment, emulsions were heat treated to tem-peratures up to 95 �C and measured also after 1 week storage. Forpartially gelatinized emulsions stored overnight, there was a ten-dency of increased lipase activity with prior amylase treatment(Fig. 7). However, this was only significant when heated to75 �C (p < 0.05). As seen in Fig. 7, the lipase activity has a mini-mum at 75 �C, then increasing again with further heating. As seenfrom Fig. 2, this could be explained by further increasing starchswelling. DSC (Fig. 1) is detecting endothermal transitions of

0

20

40

60

80

100

120

140

0 20 40 60 80 100

Lipa

se a

ctiv

ity (%

)

Temperature (°C)

Fig. 7. Lipase activity of heated emulsions stored over night without (filled circles)and with (filled squares) amylase treatment prior to the experiment, and emulsionsstored 1 week before amylase and lipase treatment (empty squares), averages andstandard deviations, n = 3.

starch during gelatinization, including loss of molecular and crys-talline order, which requires initial granular swelling, althoughsubsequent further swelling cannot be detected. The lipase activ-ity minimum at 75 �C then is interpreted as the starch being gela-tinized enough to create a rather dense layer completely coveringthe oil drop. Upon further heating, starch continue to swell, andeven though the starch layer is then thicker (Fig. 2), it will be lessdense, enabling lipase to again more easily reach the oil interface.

After 1 week storage the lipase activity after amylase treatmentare similar to the fresh sample for emulsions heated up to 75 �C.When heated to higher temperatures and then stored, the lipaseactivity of the emulsions tend to be decreased. This may be a resultof amylopectin recrystallizing at the interface. Amylopectin recrys-tallization can reduce starch digestibility even without the forma-tion of resistant starch [29]. The starch barrier will then be lessaccessible for amylase digestion and lipase transport due to theincreased structure formed.

4. Conclusions

Heating starch granule stabilized Pickering emulsions resultedin starch swelling with the starch still attached to the oil dropinterface [2,19]. It was here shown that the internal oil drop sizeremained intact although the measured emulsion drop sizeincreased upon heating due to swelling of the starch layer. The rateof gelatinization increased with temperature as expected [22].Heating times up to 10 min only had impact on starch swellingat the lower temperatures studied. Subsequent time/temperaturetreatment in order to recrystallize amylopectin [27,28] did notaffect the drop size distribution.

Starch swelling around the drop interface caused the barrier tobetter cover the oil and protect it from lipolysis [2]. Heating todifferent temperatures had clear effect on the starch barrier prop-erties. The lipase activity tended to reach a minimum for emulsionsheated to 75 �C. It was hypothesized that at this point, the starchhad gelatinized enough to create a rather dense layer protectingthe oil drop but not swelled further into a more loose structuremore accessible for the lipase to penetrate. This was supportedby micrographs and lipolysis results. Previous studies have demon-strated the possibility to gelatinize starch on emulsion drop sur-faces [2,19], however, novel effects of temperature during heattreatment and extent of swelling were obtained. Increased starchconcentrations caused more dense barriers although the lipaseactivity was reduced by more than 50% even at moderate concen-trations (>150 mg/ml oil). Furthermore, the lipolysis resultsindicated that amylopectin recrystallization reduced starchdigestibility upon storage as expected [29]. Thereby the barrierwould be less accessible for lipase transport.

These results clearly show interesting potential to create novelencapsulations. The ability to use starch granules to stabilizePickering emulsions and then apply heat to provide a controlledstarch barrier that better encapsulates the oil phase was demon-strated. This system can be used for encapsulations within the oilphase of emulsions in order to control release or to protect sensi-tive components. To obtain full understanding and control of thebarrier, further focus may be drawn to effects of fine parameteradjustments and to in vitro and in vivo digestion of the system.Since the stabilizing component is starch, it may find use in oralformulations and foods.

Acknowledgments

Anna Timgren, Hélène Trottin and Diana Marku are greatlyacknowledged for their help with practical experiments. The study

188 M. Sjöö et al. / Journal of Colloid and Interface Science 450 (2015) 182–188

was partially supported by the Lund University Antidiabetic FoodCentre, which is a VINNOVA VINN Excellence Centre.

References

[1] D.J. McClements, Y. Li, Adv. Colloid Interface Sci. 159 (2010) 213–228.[2] A. Marefati, M. Rayner, A. Timgren, P. Dejmek, M. Sjöö, Colloids Surf. A:

Physicochem. Eng. Aspects 436 (2013) 512–520.[3] B.S. Chu, G.T. Rich, M.J. Ridout, R.M. Faulks, M.S.J. Wickham, P.J. Wilde,

Langmuir 25 (2009) 9352–9360.[4] S. Emek, R. Köhnke, A. Holm, A. Szilagyi, H.-E. Åkerlund, P.-Å. Albertsson, C.

Erlanson-Albertsson, Prep. Bioch. Biotech. (2010).[5] D.J. McClements, E.A. Decker, Y. Park, Crit. Rev. Food Sci. Nutr. 49 (2009) 48–67.[6] M.V. Tzoumaki, T. Moschakis, E. Scholten, C.G. Biliaderis, Food Funct. 4 (2013)

121–129.[7] J. Frelichowska, M.-A. Bolzinger, Y. Chevalier, J. Colloid Interface Sci. 351 (2010)

348–356.[8] S. Tcholakova, N.D. Denkov, A. Lips, Phys. Chem. Chem. Phys. 10 (2008) 1608–

1627.[9] S.U. Pickering, J. Chem. Soc. 91 (1907) 2001.

[10] W. Ramsden, Proc. R. Soc. London 72 (1903) 156–164.[11] B.P. Binks, Curr. Opin. Colloid Interface Sci. 7 (2002) 21–41.[12] R. Aveyard, B.P. Binks, J.H. Clint, Adv. Colloid Interface Sci. 100–102 (2003)

503–546.[13] M. Rayner, D. Marku, M. Eriksson, M. Sjöö, P. Dejmek, M. Wahlgren, Colloids

Surf. A: Physicochem. Eng. Aspects (2014).[14] M. Wahlgren, J. Engblom, M. Sjöö, M. Rayner, Curr. Pharma. Biotechnol. 14

(2013) 1222–1234.

[15] P. Dejmek, M. Rayner, M. Sjöö, A. Timgren, New particle stabilized emulsionsand foams SE-1051328-1 PCT/SE2011/051522, 2010.

[16] D. Marku, M. Wahlgren, M. Rayner, M. Sjoo, A. Timgren, Int. J. Pharma. 428(2012) 1–7.

[17] M. Matos, A. Timgren, M. Sjöö, P. Dejmek, M. Rayner, Colloids Surf. A:Physicochem. Eng. Aspects 423 (2013) 147–153.

[18] M. Rayner, A. Timgren, M. Sjöö, P. Dejmek, J. Sci. Food Agric. 92 (2012) 1841–1847.

[19] A. Timgren, M. Rayner, M. Sjöö, P. Dejmek, Proc. Food Sci. 1 (2011) 95–103.[20] M.C. Sweedman, M.J. Tizzotti, C. Schafer, R.G. Gilbert, Carbohydr. Polym. 92

(2013) 905–920.[21] J.-L. Jane, S. Leas, H. Zobel, J.F. Roby, Starch – Stärke 46 (1994) 121–129.[22] W.A. Atwell, L.F. Hood, D.R. Lineback, E. Varriano-Marston, H.F. Zobel, The

terminology and methodology associated with basic starch phenomena, CerealFoods World 3 (1988) 306–311.

[23] N. Thomas, R. Holm, T. Rades, A. Mullertz, The AAPS J. 14 (2012) 860–871.[24] M. Rayner, M. Sjöö, A. Timgren, P. Dejmek, Faraday Discuss. 158 (2012) 139.[25] P.A. Albertsson, R. Kohnke, S.C. Emek, J. Mei, J.F. Rehfeld, H.E. Akerlund, C.

Erlanson-Albertsson, Biochem. J. 401 (2007) 727–733.[26] A. Timgren, M. Rayner, P. Dejmek, D. Marku, M. Sjöö, Food Sci. Nutr. 1 (2013)

157–171.[27] L. Elfstrand, A.-C. Eliasson, M. Jönsson, M. Reslow, B. Thelin, M. Wahlgren,

Carbohydr. Polym. 68 (2007) 568–576.[28] L. Elfstrand, A.-C. Eliasson, M. Wahlgren, Carbohydr. Polym. 75 (2009) 157–

165.[29] M.A. Leeman, M.E. Karlsson, A.-C. Eliasson, I.M.E. Björck, Carbohydr. Polym. 65

(2006) 306–313.[30] P.J. Butterworth, F.J. Warren, P.R. Ellis, Starch – Stärke 63 (2011) 395–405.[31] L.F. Harthoorn, Int. Dairy J. 18 (2008) 879–883.