Subscriber access provided by MT ROYAL COLLEGE
Article
DETERMINATION OF FORMULATION CONDITIONS ALLOWING DOUBLE EMULSIONS STABILIZED BY PGPR AND SODIUM CASEINATE TO BE USED AS CAPSULES Maxime Nollet, Eric Laurichesse, Samantha Besse, Olivier Soubabere, and Véronique Schmitt Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b04085 • Publication Date (Web): 06 Feb 2018 Downloaded from http://pubs.acs.org on February 10, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
DETERMINATION OF FORMULATION CONDITIONS ALLOWING DOUBLE EMULSIONS STABILIZED BY PGPR AND SODIUM CASEINATE TO BE USED AS CAPSULES Maxime Nollet1,2, Eric Laurichesse1, Samantha Besse3, Olivier Soubabère3 and Véronique Schmitt1,* 1
Université de Bordeaux, Centre de Recherche Paul Pascal, CNRS UMR 5031 115 Av. A.
Schweitzer, 33600 Pessac, France 2 3
Present address: Gattefossé 36 Chemin de Genas, 69800 Saint-Priest, France Natural Plant Protection, Arysta LifeScience’s group, Parc d’activités Pau-Pyrénées, 35
avenue Léon Blum, 64000 Pau, France *
corresponding author e-mail:
[email protected]
Abstract Water-in-oil-in-water
(W1/O/W2)
double
emulsions
stabilized
by
Polyglycerol
polyricinoleate (PGPR), a lipophilic food grade small polymer and sodium caseinate, a hydrophilic milk protein were developed to encapsulate Vitamin B12, a model hydrophilic substance easy to titrate. Using rheology, sensitive to drop size evolution and water fluxes, static light scattering and microscopy both giving the evolution of drops' size and Vitamin B12 titration assessing the encapsulation, we were able to detect independently, the double emulsion drop size, the encapsulation loss and the flux of water as a function of time. By differentiating the PGPR required to cover the W1-droplets' surface from PGPR in excess in the oil phase, we built a PGPR—inner droplet volume fraction diagram highlighting the domains where the double emulsion is stable towards encapsulation and/or water fluxes. We demonstrated the key-role played by non-adsorbed PGPR concentration in the intermediate sunflower oil phase on the emulsion stability while, surprisingly, the inner droplet volume fraction had no effect on the emulsion stability. At low PGPR concentration, a release of
ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 32
Vitamin B12 was observed and the leakage mechanism of coalescence between droplets and oil-water interface of the oily drops (also called globules hereafter), was identified using confocal microscopy. For high enough PGPR content, the emulsions were stable and may therefore serve as efficient capsules without need of an additional gelling, thickening, complexion or interface rigidifying agent. We generalized these results with the encapsulation of an insecticide: Cydia pomonella Granulovirus used in organic arboriculture.
Key words: Water-in-oil-in-water (W/O/W) double emulsions, encapsulation, stability, coalescence, rheology, food grade capsules, Vitamin B12, Cydia pomonella Granulovirus insecticide
1. Introduction Double emulsions made of reverse emulsion i.e. water (W1)-in-oil (O) emulsion dispersed in a continuous aqueous solution (W2), are the topic of a large research area for a long period of time. For a recent review, the reader can refer to
[1]
and references therein. Multiple
emulsions offer a possible route to protect hydrophilic substances from an aqueous environment. This makes them interesting in food industry for nutriment encapsulation like magnesium[2], vitamins[3,4], pro-biotics[5], or ω3- or n3- fatty acid prone to oxidation[6-11]. An interesting application also consists in "low fat content"[12,13] emulsions, since a part of the oil is replaced by water without texture modification. In pharmaceutical industry, double emulsions are often used as final products to encapsulate various drugs of interest as peptides[14-17], hormones[18], steroids or antiseptic[19] or as a way to produce microspheres[2022]
. Microspheres or microcapsules are produced by evaporation of the intermediate phase thus
requiring the use of a solvent. Double emulsions also exhibit the possibility to simultaneously encapsulate both oil and water compounds.
ACS Paragon Plus Environment
Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
Double emulsions exhibit very high encapsulation efficiency or yield ε defined as the proportion of the active ingredient located in the emulsion: in W1 for hydrophilic species or in O for lipophilic ones. Indeed ε belongs usually to the range 90-100%. However this encapsulation yield may decrease over time limiting double emulsion's interest for applications. The main issue about double emulsions is therefore their stability under storage. Several mechanisms leading to a decrease of the encapsulation yield of a hydrophilic species incorporated in the innermost water droplets have been described in literature
[23-27]
. The
encapsulated species may diffuse through the intermediate oily phase (ripening) or permeate through the stabilizer layers, without film rupturing, towards the continuous phase, leading to a loss of the encapsulation without altering the compartmentalized emulsion structure[19,28,29]. Alternatively inner droplets may coalesce against the surface of globule containing them. In this specific case, the initial double emulsion transforms into a simple one, widely modifying the system structure [23,27]. A difference of osmotic pressure in the two aqueous phases W1 and W2 also induces diffusion of water. If ∆π = πint – πext is positive (meaning a higher osmotic pressure πint in W1 than πext in W2), then water diffuses from the continuous phase towards the inner water droplets that swell. This may lead to the droplets burst[30] or not[31] depending on the interface resistance. On the other hand, if ∆π is negative, then water droplets shrink by emptying of their content. Several solutions have been proposed to reduce these instabilities. Very often salt has been added in the inner aqueous phase in order to decrease ripening
[32,33]
and decrease the W1
solubility in the oil phase[34]. This strategy is very efficient to reduce or stop ripening. Addition of either a gelling agent, in the inner W1 water phase[26], or a thickener in the outer W2 water phase[35-37] have also been proposed without warranting successful stability. Complexion agents in W1 can be added; for example caseinate for Ca2+, to increase the size of the encapsulated species and slow down the ripening
[38,39]
ACS Paragon Plus Environment
. Formation of complexes, for
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
example chitosan/alginate, at the interface rigidifies the globule surface
Page 4 of 32
[18,40]
. Addition of a
pressure regulator in W2 allows equilibrating πint and πext thus preventing burst or implosion. Despite the very abundant literature and the large know-how, it remains an important issue to determine and circumvent the destabilization mechanisms in order to open the door to applications. In the present paper, we propose a kinetic study based on the determination of several indicators: drops' size, emulsion elasticity, encapsulation yield, water flux, emulsion structure to determine domains where double emulsions stabilized by widespread stabilizers are efficient capsules without additional agent. We examine more carefully the influence of two parameters: the amount of lipophilic surfactant and inner droplets volume fraction on the long term stability, keeping all the other parameters constant: inner droplets' size, oily drops' volume fraction and size, concentration of hydrophilic stabilizer. If not stable, the aim is to elucidate the destabilization mechanisms. To do so, we chose to study a common system made of sunflower oil, polyglycerol polyricinoleate (PGPR) as lipophilic surfactant and sodium caseinate as hydrophilic stabilizer and we encapsulated vitamin B12 as a model of a drug. We also tested the generality of the obtained results using an insecticide used in organic farming.
2. Materials and methods
Chemicals Sunflower oil (density equal to 0.92 g.cm-3 at 20°C), sodium caseinate salt from bovine milk (CAS Number 9005-46-3 molar mass of ≈20 kDa), vitamin B12 (α-(5,6-dimethylbenzimidazolyl)cobamidcyanide) (CAS Number 68-19-9 molar mass 1 355,37 g/mol) and sodium chloride (NaCl) were purchased from Sigma-Aldrich. Polyglycerol polyricinoleate
ACS Paragon Plus Environment
Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
(PGPR) with a molar mass of 3000 g. mol-1 (data from the supplier) was provided by Paalsgard. Sodium azide was purchased from Merck and the KLD was provided by Natural Plant Protection. All the reagents were used without further purification, and Milli-Q water was used in all emulsions preparation.
Drop size measurement: The emulsion size distributions were measured using a Malvern Mastersizer MS2000 granulometer. For the W1/O emulsion, the size distribution was deduced from the angle dependent-scattered intensity using the Mie theory. The refractive indices of both the dispersed brine phase (1.336) and the continuous sunflower oil (1.476) have been measured using an Abbé refractometer. For the multiple emulsions, as the oily globules are no optically homogeneous, the simplified Fraunhofer theory was applied. We checked in a previous study[31,41] that this approximation can be used reliably for globules larger than 5µm. For both types of emulsions W1/O and W1/O/W2, the volume average diameter d and the polydispersity P, are defined by equation (1):
∑N D d= ∑N D i
4 i
1 P= Dm
i
i
3 i
i
∑N D D −D ∑N D 3 i
i
m
i
i
3 i
i
(1)
i
where Ni is the total number of droplets with diameter Di. Dm is the median diameter, i.e. the diameter for which the cumulative undersized volume fraction is equal to 50%. In the following, the inner droplets' diameter is referred to as dd and dG is the oil globules' diameter. Pd and PG are the polydispersity indices of the reverse and double emulsions respectively. The size measurements were checked qualitatively by means of an optical microscope (Zeiss, Axiovert X100).
Preparation of double emulsions:
ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Double water-in-oil-in-water emulsions were prepared following a two step procedure as described in reference [31] and specified hereafter. a) First step: The reverse emulsion was prepared by manually incorporating the W1 aqueous solution composed of 40 mM of NaCl and 0.03 wt% of vitamin B12 in the oily phase O made of sunflower oil and PGPR. The amount of PGPR has been varied from 1.64 to 6.28 wt % with respect to the oil phase. W1 (O resp) represented 40% (60% resp) in weight of the reverse emulsion (mass fraction of inner water ϕ =40%). This crude polydisperse reverse emulsion was then stirred using an Ultra-turrax® T25 rotor-stator homogenizer (Janke& Kunkel, IKA Labortechnik) equipped with a shaft (S25 KV-25F) and operating at 24000 rpm for 30 s. The resulting reverse emulsion was characterized by a mean diameter dd equal to 2 µm and a narrow drop size distribution (Pd =24%) whatever the amount of PGPR in the oil phase. Owing to the low polydispersity the surface averaged diameter also called Sauter diameter is also equal to 2 µm. We observed that stirring time had no effect on the obtained emulsion. The composition of this reverse emulsion could then be tuned by dilution either in pure oil or in oil containing PGPR to adjust both the final concentration of PGPR and water droplet volume fraction ϕ at the desired values. b) Second step: The double emulsion was prepared by using the reverse emulsion as the dispersed phase. It was progressively incorporated in an aqueous phase containing sodium caseinate at 12 wt% under manual stirring. The mass fraction of the direct emulsion was set to ϕ = 70 wt% (so that its volume fraction ϕ = 71.1%). No pressure regulator was used in the
continuous W2 phase. This crude double emulsion composed of globules containing themselves 2µm droplets was then sheared using a Couette cell (Ademtech) to reduce both the globule mean size and size distribution width. The applied shear was varied from 523 to 5250 s-1. The obtained globules size dG and polydisersity index PG defined in Eq. 1 are listed in Table 1 and an example of the reverse and double emulsions is given in Figure 1.
ACS Paragon Plus Environment
Page 6 of 32
Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
For long term experiments, an anti bacteriological agent was added. For laboratory use, sodium azide was added at a concentration of 0.03 wt% with respect to the continuous aqueous phase. As this agent is toxic and not adapted for the targeted application, we also tested the addition of the so called KLD at 2 wt%, a 12-residue self-assembling peptide authorized in food and organic farming [42]. We checked that the addition of these two agents did not modify the emulsions characteristics and behaviors.
Interfacial tension: For interfacial tension values larger than 2 mN.m-1, the pendant-drop method was used. A drop of the aqueous phase was formed at the tip of a needle plunged in the oily phase. The interfacial tension was deduced from the equilibrium axi-symmetric drop shape using the Laplace equation. For lower values, a spinning drop device was used in order to perform the interfacial measurements. The two phases were then put into contact to allow equilibration for 24 hours. Then an oily drop, containing various amounts of PGPR (less dense fluid), was deposited in the cylindrical tube containing water and NaCl (more dense fluid). A fast rotation (2000 to 6000 rpm) of the cylinder induced the deformation of the oily drop. At equilibrium, centrifuge forces, that tend to elongate the drop, and interfacial forces, that tend to minimize the interfacial area and hence resist the deformation, compensate. In all cases, the spinning drop length was at least four times larger than its diameter. The values of the interfacial tension γ could then be deduced using the classical relation: =
(2)
where Δ is the density mismatch between the two liquids, !" is the rotation speed. Therefore the interfacial tension is given by the determination of the cylindrical diameter #$ .
ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 32
Rheology: The experiments were performed using an AR1000 TA rheometer equipped with parallel plates geometry (40 mm) with a gap of 1 mm. Despite the fact that the strain and stress are not constant over the whole sheared volume, this type of geometry was preferred in order to avoid problems related to confinement that may happen in a cone and plate geometry owing to the cone truncation. The plate surfaces were made rough by sand blasting in order to avoid any slip at the walls. A solvent trap prevented water evaporation from the samples during the measurements. We checked that the results were geometry independent. The elastic G' and loss G" moduli were measured in the oscillatory mode and in the linear visco-elastic regime (LVER). As no frequency dependence has been observed for both moduli, the frequency was set constant and equal to 1 Hz throughout the study. In a previous paper[31] we established that, the elastic modulus G' of a double emulsion is directly related to the effective globule volume fraction through (supporting information S1): %& =
'( )
0
*ϕ +,, − ϕ. /
(3)
where R is the emulsion drop radius (R=dG/2). This relation is a consequence of the powerlaw expression proposed by Trappe et al. in their unified description of colloidal gels[43]. ϕ +,, is the effective globule volume fraction that may differ from ϕ = 71.1% resulting from the composition and the preparation if some water exchanged between the two compartments. The expression of G' is system-dependent i.e. the values of the constant A0 and ϕ. depend on the attractive interactions between colloids. We showed earlier[31] that for the studied system A0 = 0.35 Pa.m and ϕ. = 0.43 independently of the globule content details i.e. of the number of inner droplets. Water droplets only intervene through their contribution to the global value of ϕ +,, . An observed variation of the emulsion elasticity therefore results either from an
ACS Paragon Plus Environment
Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
evolution of the oily drop size, R, that may be detected by light scattering or from a variation of the oil drop volume fraction ϕ +,, , owing, for example, from water exchanges or from an encapsulated species release. Note that the high power law in Eq. 3 makes rheology a very sensitive tool to detect an evolution of ϕ +,, even if it does not correspond to a detectable difference in R. As a consequence ϕ +,, can be assessed as a function of time by simply measuring the double emulsion elasticity (via rheology) and drop radius (via static light scattering) using the relation: ϕ +,, = 0.43 + 6
&7 :/0
.. 8
9
(4)
Encapsulation efficiency (ε) Vitamin B12 or α-(5,6-dimethylbenzimidazolyl)cobamidcyanide is a water soluble vitamin having a significant role in the normal functioning of the brain and nervous system, and for the formation of blood. Vitamin B12 exhibits absorbance peaks at 360 nm and 550 nm allowing easily measuring its concentration in the external W2 aqueous phase after establishing a calibration curve i.e. out of the capsules [31]. The low limit value of detection is estimated around 0.001g of vitamin B12 per g of aqueous phase[31]. To determine encapsulation, 5 g of the double emulsions were first diluted with 1.5 g of 40 mM NaCl brine and slightly centrifuged (≤400 g) in order to separate the globules from the continuous phase without deteriorating the double emulsion. If necessary, the continuous phase was taken and filtered to avoid any un-creamed oil globule or excess of un-adsorbed caseinate to pollute the sample. The encapsulation efficiency is then deduced using the following relation:
?@AB ?C DEB=>EA F: EA BGH $=IJ@KHJ B?B=K =>?@AB ?C DEB=>EA F:
=1−
=>?@AB ?C DEB=>EA F: ?@B ?C BGH $=IJ@KHJ B?B=K =>?@AB ?C DEB=>EA F:
ACS Paragon Plus Environment
(5)
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 32
Just after emulsion preparation ε is equal to 90±3%. We checked that relation 5 can be applied by measuring vitamin B12 in the aqueous droplets by separating the continuous phase from the double emulsion before applying a high enough centrifugation to destroy the double emulsion and to induce the release of the innermost aqueous phase. The two measurements gave consistent values of encapsulation. After this test, ε was just determined by measuring the vitamin B12 concentration in the outermost aqueous phase. Cydia pomonella Granulovirus (CpGV) is a natural insecticide used in Integrated Pest Management (IPM) or organic agriculture to control the codling moth (Cydia pomonella), [44] a Tortricidae Lepidoptera devastating numerous fruit trees such as apple, pear and quince. The virus is produced by an in vivo multiplication by infected larvae as described in ref. [45]. After their death, larvae are crushed and purified. A purified suspension of 7.1014 viral entities per liter of CpGV in water was provided by Natural Plant Protection. This suspension has been diluted with brine to 1012 viral entity per liter prior to emulsification. Milli-Q water was used in all emulsions preparation. The determination of virus encapsulation has been derived by counting the virus under observation with an optical microscope equipped with a dark field. The results 94±3% were comparable with those for vitamin B12 [46].
Confocal microscopy: In order to better identify the destabilization mechanisms and to discriminate between ripening and coalescence, emulsions were observed by means of a confocal microscope equipped with filters. 0.02 wt% of Rhodamine B was previously added to W1 and replaced vitamin
B12.
Rhodamine
B
also
called
[9-(2-carboxyphenyl)-6-diethylamino-3-
xanthenylidene]-diethylammonium chloride is a water soluble dye with a molar mass equal to
ACS Paragon Plus Environment
Page 11 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
479.01 g.mol-1 often used in microbiology. Its maximum excitation and emission wavelengths are 544 and 610 m respectively.
3. Results and discussion 3.1 Characterization of the emulsion Double emulsions W1/O/W2 made of 40 mM NaCl/sunflower oil/12wt% of sodium caseinate in water were prepared as described in the Materials and Methods §. Their composition and characteristics are described in Table 1. Table 1: Sum up of the double emulsion characteristics: composition, applied shear rate in the second emulsification step and resulting oily drop size distribution (dG is the volume average globule diameter and PG the polydispersity index defined in Eq. 1). MO N
Total
Non adsorbed
PGPR (wt%)
PGPR (wt%)
Shear rate (s-1)
dG
PG
(µm)
40%
1.64
0.36
5250
14.4
0.25
20%
1.00
0.36
3140
13.2
0.27
10%
0.68
0.36
1050
18.7
0.30
5%
0.52
0.36
523
19.0
0.22
40%
2.28
1.0
5250
12.2
0.25
20%
1.64
1.0
4200
9.70
0.20
10%
1.32
1.0
1050
16.8
0.27
5%
1.16
1.0
523
17.1
0.35
40%
3.78
2.5
5250
14.0
0.25
20%
3.14
2.5
3140
8.60
0.20
10%
2.82
2.5
1050
10.8
0.27
5%
2.66
2.5
523
10.6
0.40
40%
6.28
5.0
4200
9.20
0.27
20%
5.64
5.0
3140
5.80
0.45
10%
5.32
5.0
1050
4.20
0.40
ACS Paragon Plus Environment
Langmuir
5%
5.16
5.0
523
5.20
0.30
ϕ is the weight proportion of W1 in the reverse emulsion, ϕ is set to 70 wt%, the amount of PGPR is given in weight % with respect to the oily phase O. The inner water droplets have a constant size dd= 2µm and a polydispersity of 0.24. PGPR concentration values in blue and in italic are calculated (see text). Emulsions with ϕ P 40% were obtained by dilution of the mother ϕ = 40% reverse emulsion.
a)
b)
volume (%)
15
10
5
0 0.01
0.1
1
10
100
1000
Diameter (µm)
c)
25
d) 20
volume (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 32
15 10 5 0 0.01
0.1
1
10
Diameter (µm)
100
1000
Figure 1: Preparation of the double emulsion. In a first step the reverse water-in-oil (a and b) is prepared before becoming the dispersed phase of the double emulsion (c and d). The double emulsion composition was as following: ϕ =40%, ϕ =70%, 2.28wt% of PGPR with respect to the sunflower oil and 12wt% of sodium caseinate with respect to the continuous aqueous phase. The emulsions were diluted before observation by means of an optical microscope (a and c: reverse and double emulsions micrographs respectively) and before static light scattering experiments (b and d: reverse and double emulsions drop size distributions respectively). Adapted from ref.[31]
In order to get better insight into the PGPR distribution between the inverse emulsion droplets and the oil, we first studied the brine and the sunflower interface in presence of various amounts of PGPR and various concentrations of NaCl in water (from 0.04 to 0.1M).
ACS Paragon Plus Environment
Page 13 of 32
In this range of NaCl concentration, no evolution could be measured. An example is given in Figure 2. 14
Interfacial tension (mN/m)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
12 10 8 6 4 2 0 -10
-8
-6
-4
-2
0
Ln (c)
Figure 2: Interfacial tension between brine (NaCl 0.1M) and sunflower oil containing increasing amounts of PGPR. The PGPR is expressed in mol/L with respect to the oil phase. The experiments were performed with samples prepared in weight fractions. The weight fraction was then converted into mol/L taking a molar mass of 3000 g/mol and an oil density of 0.92. The cmc was obtained for 1wt% (≈3.1 10-3 mol/L). The same data with c expressed in mass fractions with respect to the oil phase is plotted in supporting information S2.
From such a curve, the critical micellar concentration (cmc) defined as the concentration above which the interfacial tension is constant can be deduced (1wt%) as well as the surface concentration Γ using Gibbs equation: Γ = −
:
7S
∗
dγ dln(c)
=1.67 10-6 mole.m-2 leading to a
surface per polar head of 100 Å2. It has been checked that the PGPR adsorption at the oil/water interface is reversible assessing the validity of Gibbs equation (Supporting information S3). As the volume and surface-averaged droplet size dd was constant and equal to 2 µm, for each emulsion prepared at ϕ =40%, the amount of non-adsorbed PGPR (free in the oil phase O) could be calculated (see Table 1). Then, the reverse emulsion was diluted with oil to tune ϕ at the corresponding PGPR concentration, to keep the free amount of PGPR constant. We could therefore prepare double emulsions with variable number of inner droplets nd keeping the amount of lipophilic stabilizer constant in the intermediate oil phase
ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 32
(see Table 1). As a consequence we were able to determine the influence on the stability of each of these two parameters, keeping everything else constant.
3.2 Stability diagram Once prepared, double emulsions were stored at room temperature and, at regular time intervals, a small amount of them was taken to observe the double emulsions by optical microscopy and to measure the drop size distribution (dG, P), the elastic modulus G' and the encapsulation ε. Three types of emulsion evolutions (that will be described in detail) with time could be distinguished: they are represented by three different symbols. All the obtained results are gathered in the following graphs corresponding to (PGPR— ϕ ) stability diagrams (Figure 3).
a)
b)
Figure 3: Stability diagrams of double emulsions stabilized by various amounts of PGPR and a fixed amount (12 wt% with respect to the outer continuous phase) of sodium caseinate. The oil globule fraction was fixed to 70 wt% and the inner droplet weight fraction was varied up to 40%. The diagram is plotted a) as a function of the total PGPR amount or b) as a function of the free PGPR amount in the oil phase. Experimental data are plotted as symbols: red squares G'R that is to say the effective globule volume fraction ϕ +,, and ε decrease as a function of time, blue discs G'R that is to say ϕ +,, and ε are constant with time and green +,, triangles G'R or in other words ϕ decreases with time while ε is constant. The lines are guides to the eyes as the limits have not been determined precisely.
ACS Paragon Plus Environment
Page 15 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
It is worth mentioning that the simple direct emulsion O/W2, that is to say if no inverse emulsion is incorporated in the globules ϕ = 0, was stable whatever the PGPR concentration in the oil. From these diagrams three different domains noted 1 to 3 can be distinguished. Note that when plotted as a function of non-adsorbed PGPR concentration, in O, these domains are independent of the inner droplet fraction meaning that the inner droplet volume fraction ϕ had no influence on the double emulsion stability.
3.2.1 Domain (1) – Low free PGPR concentration In the 1st domain (1), where the symbols are red squares, the amount of free PGPR is equal to 0.36 wt% – that is to say lower than the determined cmc in sunflower oil – the oil drop size as well as the emulsion compartmentalized structure did not evolve. However both the elasticity and consequently the deduced effective globule volume fraction ϕ +,, and the encapsulation ε decreased with time (see Figures 4 and 5).
a)
b)
Figure 4: Optical micrographs of the double emulsions with ϕ = 40% and 0.36wt% of non adsorbed PGPR in oil, a) just after preparation and b) after 36 days.
ACS Paragon Plus Environment
10
-2
10
-3
Page 16 of 32
10-4 40% 20% 10% 5%
10
-5
0
5
10
15
20
25
30
35
40
Time (days)
a) 80
v eff
(%)
70
φG
60
50
40
40% 20% 10% 5%
0
10 20 Time (days)
30
40
b) 100 80
ε (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
G' R(Pa.m)
Langmuir
60 40 40% 20% 10% 5%
20 0
c)
0
5
10
15 20 25 Time (days)
30
35
40
Figure 5: Kinetic evolution of a) G'R where G' was measured through rheology and R by granulometry, b) ϕ +,, deduced by means of relation 4, the dashed line corresponds to the expected value (for clarity reasons the error bars are only drawn on the last points) and the small colored dashed lines at the right of the graph correspond to the case where all the innermost water exits and c) the encapsulation efficiency ε for emulsions belonging to domain 1: 0.36 wt% of free PGPR and ϕ varies from 5 to 40%.
ACS Paragon Plus Environment
Page 17 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
In order to get better insight into the destabilization mechanisms occurring in the first domain, it was necessary to visualize inner droplets to count them and assess their possible size evolution. To make their observation possible using a confocal microscope, we replaced vitamin B12 by a small amount of Rhodamine B, a hydrophilic fluorescent molecule, in the W1 aqueous phase (keeping the other components constant). Its concentration was set to 0.02wt% with respect to the aqueous phase. Also, as the stability diagram is independent on the droplet volume fraction, we fixed it at ϕ =10%, a value low enough to make the droplets well distinguishable. The emulsion was observed at various time intervals over a period of three months by means of a confocal microscope. Two pictures, taken just after emulsion preparation and after 88 days, are reported in Figure 6.
Figure 6: Emulsion belonging to domain 1 with initially ϕ =10% a) after preparation and b) after 88 days.
Initially the fluorescence is located in the capsules and more specifically in the inner droplets confirming the encapsulation ability of the double emulsion. Then, as time passes, the continuous phase became fluorescent as well, confirming the encapsulation loss already observed for vitamin B12. It also appears that the number of inner droplets has reduced
ACS Paragon Plus Environment
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 32
drastically between the two pictures while both the globules and the droplets' sizes have not varied significantly enough to be visible. The observations were the same with the encapsulated virus Cydia pomonella Granulovirus
[46]
. Additionally, whatever the
encapsulated species, we did not observe any noticeable evolution of the amount of oil globules showing that no significant globule-globule coalescence occurred. From these observations, one can conclude that the encapsulated species release is due to coalescence between droplets and the oil globule surface. Note that in this PGPR concentration domain, the amount of surfactant in the oil phase (non-adsorbed) is 0.36wt% lower than the determined cmc. We therefore hypothesize that this instability results from an insufficient amount of lipophilic surfactant. This domain can therefore be called "PGPR-poor regime". Note that the instability disappears when the free PGPR concentration becomes enough that is to say when it reaches its cmc. From a statistical analysis of pictures similar to those reported in Figure 6, the average number of inner droplets per globule can be plotted as a function of time (see Figure 7). As it has been reported earlier[23], occurrence of droplet/globule coalescence first requires the adsorption of droplets on the inner globule surface. Such phenomenon is clearly visible in Figure 6. Adsorption probably results from van der Waals attraction between droplet and globule surfaces. This interaction energy can be estimated through: UDV = −
WXY
(6)
: ZX[Y (X]Y )
where A is the water-sunflower oil-water Hamaker constant estimated
[47,48]
(we did not take
into account the PGPR for A) to be equal to 0.6 10` . J, dd=2 µm and Dd-G is the dropletglobule distance that is estimated by addition of the sodium caseinate gyration radius in the oil that can be neglected and of the PGPR brush length considering 3 PR units: Dd-G ≈ 3 nm. This leads to Wvdw ≈ -70 kBTa, kB being Boltzmann's constant and Ta the ambient
ACS Paragon Plus Environment
Page 19 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
temperature=298K. This rough estimation shows that droplets strongly interact with the globule surface. It becomes interesting to determine the number of adsorbed droplets na per globule compared to the total number of droplet per globule nd. As the interaction energy is high, na corresponds to the droplets that have enough room to adsorb at the globule surface. We can therefore define the coverage parameter C as the proportion of globule surface covered by droplets considering that the surface space occupied by a droplet is its projected surface i.e. its equatorial surface, Seq d. Then: a=
AXJbc X
defghifb
=
AX j X 0j
6 9 = Y
l kX Y
0
6 9 X
(7)
For example, if mD =0.1, dd=2 µm and dG=16 µm, C=0.2 and becomes equal to 0.8 for mD =0.4. Whatever the considered droplet volume fraction, C is smaller than 1 meaning that there is enough room at the globule surface for all the droplets to be adsorbed (proportion of adsorbed droplets close to 1). Therefore the number of adsorbed droplets na can be approximated by nd. This feature is in agreement with Figure 6. For each droplet/globule coalescence event, the number of droplets is reduced by one unit. This means that the coalescence over a time interval dt can be described by: #n = −!n= #o where ω is the frequency of a coalescence event and na=nd. The straightforward solution of this equation is given by: n (o) = n (0)p `B
(8)
Using confocal microscopy, we could determine the average number of droplet per globule as a function of time n (o). To do that, we carefully scanned the globules vertically taking care to count each droplet only once. The kinetic evolution of the number of droplet per globule is plotted on Figure 7. The best fit of the experimental data to Equation 8 leads to n (0) =58.7 and ω=1.26 10-5 min-1. The coalescence frequency normalized by the initial droplet surface area is ωS0=1.81 104 min-1.m-2.
ACS Paragon Plus Environment
Langmuir
70 60 50 40
nd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 32
30 20 10 0
0
10
20
30
40 50 60 Time (days)
70
80
90 100
Figure 7: Average number of water (W1) droplets nd per globule (O) as a function of time for the emulsion containing 0.36 wt% of free PGPR and ϕ =10%. The symbols are the experimental data and the line is the best fit to equation 8.
The coalescence frequency designates a microscopic property of the liquid film made of the droplet surface, the intermediate oil phase and the globule surface. Indeed τ=1/ω is the average lifetime of the film, therefore the lower the ω, the more stable, the film. In the present case, τ is equal to 2 days and 7 hours. The comparison with other systems [23,49-51] is not easy because of the very different sizes of the droplets and globules. The initial droplet number extracted from the best fit (59±5) can be compared to the estimated value deduced from the emulsion composition. Indeed nd(0) can also be written as
n = mD 6 Y 9 . For mD =0.108 (mV =0.1), dd=2 µm and dG=16 µm, nd=55.3. Therefore the X
agreement between both values, the one estimated from the composition and the one given by the fit, is pretty good. For each coalescence event, the droplet content is released in the continuous phase. It is therefore possible to transform the experimental data of nd(t) into an encapsulation curve εcounting(t) assuming a homogeneous distribution of the encapsulated species in the droplets. These data can be compared with vitamin B12 release kinetics measured via UV-Vis spectroscopy of the continuous phase (Figure 8).
ACS Paragon Plus Environment
Page 21 of 32
100 90 80 70
ε (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
60 50 40 30 20 10 0
0
10
20
30
40
50
60
70
80
90 100
Time (days))
Figure 8: Encapsulation efficiency in the emulsion containing 0.36wt% of free PGPR and ϕ =10%. ε is obtained either by vitamin B12 titration (black triangles) or by counting the release arising from the coalescence of water droplets onto the oil globules (red squares).
The two curves superimpose, demonstrating the validity of the droplet/globule coalescence mechanism. It is worth noticing (Fig 5b and c) that at low inner droplet volume fraction (5 and 10%) although the trends for the evolution of both the effective globule volume fraction and the encapsulation rate qualitatively agrees, the agreement is not quantitative while the agreement was excellent between the encapsulations rates deduced either by counting the number of droplet and the titration. Indeed after 36 days encapsulation remains at a level close to 50% while the effective volume fraction has decreased at its minimum value. Therefore the evolution of the elastic modulus is a good indication for identifying the water flux: the evolution of G' is even macroscopically perceptible (emulsion texture evolution) however one has to take care about the quantitative value at least for low droplet volume fractions.
It is known in the literature that addition of a hydrophilic surfactant in the W2 continuous phase above its cmc induces droplet/globule coalescence[23]. We aimed at checking this effect on the present system. Therefore after the double emulsion preparation, we added, in W2, sodium dodecyl sulphate (SDS) a well known hydrophilic surfactant so that its final
ACS Paragon Plus Environment
Langmuir
concentration is equal to 10 cmc with respect to W2 and proceeded in a similar way to determine the average number of droplets per globule as a function of time and the encapsulation kinetic release. We are aware that SDS is not a surfactant that can be used in organic farming but it is interesting to assess in the laboratory, the possibility of provoking the
encapsulation (%)
release at will with a model surfactant. The results are reported in Figure 9.
100 90 80 70 60 50 40 30 20 10 0
0
10000
20000
30000
40000
50000
60000
S D S h t i w
100 90 80 70 60 50 40 30 20 10 0
S D S t u o h t i w
0
25
50
75
100
125
150
175
200
Time (min)
a)
55 50 45 40 35 30 25 20 15 10 5 0
nd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 32
0
b)
50
100 Time (mn)
150
200
Figure 9: a) Encapsulation efficiency of vitamin B12 in a double emulsion containing 0.36wt% of free PGPR and ϕ =10%, without (red diamonds) and with (black discs) SDS and b) kinetic evolution of the average number of water droplets per oil globule with addition of SDS in the outer continuous phase. The symbols are the experimental data and the line is the best fit using equation 8.
ACS Paragon Plus Environment
Page 23 of 32
Even if the outer surface is protected by a protein, addition of SDS has the same effect improving the droplet/globule coalescence and hence accelerating the kinetics of drug release, likely by weakening the protein layer. Indeed the best fit to experimental data gives ω=1.46 10-2 min-1 (the coalescence frequency normalized by the initial droplet surface area is ωS0=2.10 107 min-1.m-2) corresponding to a film lifetime of 1 hour and 8 minutes that is to say an acceleration of the kinetics by a factor close to 1000, keeping the droplets and globules sizes unchanged. This shows the importance of the system composition. Addition of a hydrophilic surfactant in the continuous aqueous phase is therefore a good way to accelerate the release at wish (boost effect).
3.2.2 Domain (2) – Free PGPR concentration close to cmc In the 2nd domain (Figure 3) where the symbols are blue circles, the amount of free PGPR is equal to 1wt% – that is to say close to its cmc in sunflower oil – no significant evolution of R, G' and ε could be observed over time and the double emulsions were stable over 9 months (Figure 10 a and b).
10
-3
G'R (Pa.m)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
10
a)
40% 20% 10% 5% -4
0
50
100 150 200 Time (days)
ACS Paragon Plus Environment
250
300
Langmuir
100 90 80 70 60
ε (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 32
50 40 30
40% 20% 10% 5%
20 10 0
b)
0
50
100
150
200
Time (days)
250
300
Figure 10: Kinetic evolution of a) G'R and b) encapsulation efficiency ε for emulsions belonging to domain 2: 1 wt% of free PGPR and ϕ varies from 5 to 40%. This means that no water or encapsulated species exchange occurred. Phases are well equilibrated and interfaces are protected enough. In this domain, restricted in PGPR content but extended with respect to ϕ , double emulsions are stable and completely leak-free capsules. This constitutes therefore a very interesting domain for encapsulation applications. Note that in this regime, no addition of gelling agents, thickeners, complexion agents, or interface rigidifying agents was required to obtain tight capsules. Double emulsions are very stable kinetically (more than nine months) and they are ideal systems that may serve as impervious capsules. The same result has been observed with the organic insecticide Cydia pomonella Granulovirus [46].
3.2.3 Domain (3) – High free PGPR concentration
In the third domain (3), corresponding to higher free PGPR concentrations (≥2.5wt%), the encapsulation efficiency ε remained constant while G'R decreased with time, meaning an outgoing water flux without transport of vitamin B12 towards the continuous phase (see
ACS Paragon Plus Environment
Page 25 of 32
Figures 11) over a period larger than nine months. The same results -decrease of the emulsion elasticity without altering the compartmentalized structure and the drops size and without loss of encapsulation yield- were obtained for Cydia pomonella Granulovirus [46] -2
10
-3
G'R (Pa.m)
10
-4
10
40% (2.5% PGPR) 20% (2.5% PGPR) 10% (2.5% PGPR) 5% (2.5% PGPR)
40% (5% PGPR) 20% (5% PGPR) 10% (5% PGPR) 5% (5% PGPR)
-5
10
0
50
100
150
200
250
300
Time (days)
a) 80
60
φG
v eff
(%)
70
40% (2.5% PGPR) 20% (2.5% PGPR) 10% (2.5% PGPR) 5% (2.5% PGPR)
50
40
0
50
100
40% (5% PGPR) 20% (5% PGPR) 10% (5% PGPR) 5% (5% PGPR)
150
200
250
300
Time (days)
b) 100 90 80 70 60
ε (%)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
50 40 30
40% (2.5% PGPR) 20% (2.5% PGPR) 10% (2.5% PGPR) 5% (2.5% PGPR)
20 10 0
c)
0
40% (5% PGPR) 20% (5% PGPR) 10% (5% PGPR) 5% (5% PGPR)
100Time (days)200
ACS Paragon Plus Environment
300
Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 26 of 32
Figure 11: Kinetic evolution of a) G'R, b) ϕ +,, deduced by means of relation 4, the dashed line corresponds to the expected value (b: the error bars have only been reported on the last points for clarity reasons) and the small colored dashed lines at the right of the graph correspond to the case where all the innermost water exits and c) the encapsulation efficiency ε for emulsions belonging to domain 3: 2.5 wt% and 5wt% of free PGPR and ϕq varies from 5 to 40%.
This water flux becomes faster as the water droplet fraction increases whereas it is independent of the free PGPR concentration. As vitamin B12 remained encapsulated and no disappearance of the inner droplets was observed, droplet/globule coalescence can be ruled out. These observations are only compatible with ripening or permeation. Ripening or permeation requires a chemical potential difference due to either the Laplace pressure of small droplets or to ions concentration difference and therefore osmotic pressure difference in the two aqueous compartments W1 and W2. The osmotic pressure in W2 comes from the dissolution of 25 ions per sodium caseinate molecule. As the concentration of sodium caseinate in W2 is equal to 12 wt% corresponding to 6 mM, the osmotic pressure P2 in W2 arises mainly due to the ions (0.15 M) and can be roughly estimated: P2=386 kPa (371 kPa from the ions and 15 kPa from the protein). In the droplets, one has to consider three contributions to the total pressure P1: the osmotic pressure from 40 mM NaCl (198 kPa) and from 0.22 mM vitamin B2 (0.55 kPa) and the Laplace pressure PL =
t )
= 4 kPa. Laplace
pressure of droplets is small compared to osmotic pressures arising from NaCl so that it can be neglected. The osmotic pressure being much larger in W2 than in W1, (P2=386 kPa > P1=202 kPa) or in other words the ions concentration c2=0.156 M being larger in W2 than the ion concentration c1=40.22 mM in W1, water tends to exit as was demonstrated in a previous paper[31]. Note that in the previous domain (intermediate PGPR concentration, domain noted 2) the pressures P1 and P2 are the same than in the third domain, but no water exchange was observed. If this is the motor for water exchange, then it seems that the presence of micellar aggregates is necessary for water diffusion in the oil phase even if an increase of PGPR by a
ACS Paragon Plus Environment
Page 27 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
factor 2 did not accentuate water transfer. As was already seen previously where an even much stronger deliberate pressure unbalance was applied[31], the capsules are vitamin B12 tight. Note that the kinetics of water exchange when micelles are present is surprisingly slow. In the absence of excess of micelles, the pressure unbalance was almost immediately compensated by water exchange. We think that in this domain, the origin and mechanism of outgoing water remains not elucidated. However, even if water flux is observed, as vitamin B12 or Cydia pomonella Granulovirus are still encapsulated, this domain remains very interesting for encapsulation applications.
Conclusions We showed the importance of free, non-adsorbed, lipophilic PGPR concentration in the middle phase of water-in-oil-in-water (W1/O/W2) double emulsions stabilized by PGPR and sodium caseinate. In contrast, in this chosen representation the inner water droplet volume fraction had no impact on emulsion stability. By distinguishing the lipophilic surfactant used to cover the droplets surface from the one dissolved in the oil phase, we could distinguish three domains as a function of PGPR concentration. Two of them, at concentrations larger than 1wt% are very interesting for applications since the encapsulation remains at the high level, reached just after emulsion preparation, for a period of at least 300 days. This is accomplished without addition of further compound. Moreover the osmotic pressure regulator in the continuous aqueous phase may be removed without provoking a decrease of the encapsulating potential. While an outgoing water flux, that does not alter encapsulation, could be evidenced via rheology for high free PGPR concentration (≥2.5wt%) no evolution at all could be detected in a narrow concentration range around the PGPR cmc (1wt%). In the PGPR-poor domain (PGPR
