Water-in-water emulsion gels stabilized by cellulose nanocrystals

32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52 .... system was presheared during 1 min at 100 s-1 before starting ...
1 downloads 3 Views 1MB Size
Subscriber access provided by READING UNIV

Interface Components: Nanoparticles, Colloids, Emulsions, Surfactants, Proteins, Polymers

Water-in-water emulsion gels stabilized by cellulose nanocrystals. Emna Ben Ayed, Remy Cochereau, Cyrille Dechance, Isabelle Capron, Taco Nicolai, and Lazhar Benyahia Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.8b01239 • Publication Date (Web): 19 May 2018 Downloaded from http://pubs.acs.org on May 19, 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.

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 20 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

Water-in-water emulsion gels stabilized by cellulose nanocrystals.

Emna Ben Ayed1,2, Remy Cochereau1, Cyrille Dechancé1, Isabelle Capron3, Taco Nicolai1, Lazhar Benyahia1 1 Le Mans Université, IMMM UMR-CNRS 6283, 72085 Le Mans cedex 9, France 2

Faculty of Science of Sfax -University of Sfax, BP 1171-3000 Sfax-Tunisia

3

UR1268 Biopolymères, Interactions et Assemblages, INRA, F-44316 Nantes, France

email: [email protected]

1

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

Abstract Particle stabilized water-in-water emulsions were prepared by mixing dextran and poly(ethylene oxide) (PEO) in water and adding cellulose nanocrystals (CNC). The CNC formed a layer at the surface of the dispersed droplets formed by the PEO rich phase. Excess CNC partitioned to the continuous dextran phase. Aggregation of CNC at different rates was induced by adding NaCl between 10 mM and 100 mM. In the presence of more than 2 g/L CNC, fast aggregation led to the formation of an emulsion gel showing no signs of creaming. Confocal laser scanning microscopy showed that the emulgels were formed by a continuous network of CNC in which the randomly distributed droplets were embedded. The gel stiffness was measured with oscillatory shear rheology and found to increase strongly with increasing CNC concentration (C). The dispersed droplets were elastically active and increased the gel stiffness at low C. However, up to C = 10 g/L the yield stress was too small to inhibit flow when the gels were tilted. At C < 2 g/L creaming was observed until the network of connected droplets became sufficiently dense to be strong enough to resist buoyancy. Keywords: Pickering emulsions; particle stabilization; cellulose whiskers; aqueous two phase; phase separation; emulgel

2

ACS Paragon Plus Environment

Page 2 of 20

Page 3 of 20 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

Introduction When two aqueous solutions containing different incompatible macromolecules are mixed at a high enough concentration, water droplets containing one of the macromolecules are formed in a continuous water solution of the other macromolecule

1, 2

. Such systems are called water-in-water

(W/W) emulsions analogous to oil-in-water (O/W) emulsions where oil droplets are formed in a continuous water phase. A major difference between W/W and O/W emulsions is that the interfacial tension (γ) of the former is orders of magnitude smaller and can be varied continuously by varying the concentration of the macromolecules in each phase 3. Another important difference is that W/W emulsions cannot be stabilized against coalescence of the dispersed droplets by surfactants, because the interface is expressed only on length scales larger than the correlation length of the macromolecules in solution. It has been shown, however, that stabilization can be achieved by adding nano- or microparticles

4-14

similarly to so-called Pickering O/W emulsions

15, 16

. If these

particles have some affinity with both types of macromolecules they will position themselves at the interface and thereby reduce the contact area between the two incompatible solutions. Even though W/W emulsions can be stabilized against coalescence, creaming or sedimentation of the dispersed phase will occur at a rate that depends on the droplet size, the density difference between the two phases and the viscosity of the continuous phase

12, 17

. This

phenomenon can be inhibited by gelling the emulsions. O/W emulsion gels, or emulgels, have been investigated in some detail and one may distinguish two limiting situations18. When the continuous phase contains a relatively high concentration of macromolecules that are crosslinked, the emulgel is a homogeneous network with dispersed droplets randomly distributed within the network. Alternatively, if the dispersed droplets bind to each other the emulgel consists of a space filling network of crosslinked droplets within a liquid continuous phase. In many cases the structure will be intermediate with crosslinked droplets in a gelled continuous phase. It would be of practical interest if the same particles that stabilize the dispersed droplets against coalescence can also be used to produce emulgels and thus stabilize against creaming or sedimentation. It was recently shown that this can be done successfully for W/W emulsions formed by mixing aqueous solutions dextran and poly(ethylene oxide) (PEO) in the presence of protein microgels19. Excess microgels in the continuous phase could be induced to aggregate and form a network by adding salt or reducing the pH, which reduces the electrostatic repulsion between the microgels. W/W emulsions formed by mixing aqueous solutions of pullulan and PEO 8 or gelatin and

3

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

PEO

14

Page 4 of 20

in the presence of clay particles were found to be stable against sedimentation of the

dispersed droplets, because aggregated clay particles attached to the droplet surface connected the droplets into a space filling network. However, the rheological properties of W/W emulgels as a function of the concentration of particles and the volume fraction of the dispersed phase has not yet been investigated. Here we present such an investigation for PEO-in-dextran emulsions in the presence of cellulose nanocrystals (CNC). CNC are rod-like negatively charged particles produced by acidic hydrolysis of cellulose20, 21. It was shown elsewhere that CNC adsorbs at the interface and effectively inhibits coalescence of dispersed PEO droplets 12. Based on free energy calculations it was concluded the preferred orientation of the particles is parallel to the droplet surface12, 14. Excess CNC was found to be almost exclusively situated in the continuous dextran phase. CNC remains dispersed in water due to strong electrostatic repulsion, but when NaCl is added electrostatic interactions are screened allowing CNC to bind together and to form a self-supporting gel at C ≥ 3 g/L

22-24

. Gels at low CNC

concentrations are very weak and flow when tilted for C < 12 g/L. The rate of CNC gelation was found to increase strongly with increasing salt or CNC concentration. We have investigated the effect of CNC gelation on the stability of the emulsions against creaming as a function of the CNC and the NaCl concentration. It will be shown that the steady state structure of the emulgels depends on the relative rate of creaming and gelation and therefore on the NaCl concentration. By varying the CNC concentration emulgels could be studied with structures that cover the whole range between a network of crosslinked PEO droplets to a continuous CNC network with embedded randomly distributed droplets. The stiffness of the emulgels will be compared to that of the gelled continuous phase so that the contribution of the dispersed phase can be assessed.

Experimental The dextran (5 × 105 g/mol), PEO (2 × 105 g/mol), NaCl and Nile blue A were purchased from Sigma-Aldrich. Dextran was used as it was received but PEO contained a small amount of silica particles that were removed by centrifugation at 5 × 105 g. Cellulose nanocrystals (CAS Number: 9004-34-6) were purchased from the university of Maine process development center (US). The powder was dispersed in Millipore water by ultrasonification using a sonicator from Bioblock Scientific. The molar mass (Mw), radius of gyration (Rg) and hydrodynamic radius (Rh) of the CNC were determined using static and dynamic light scattering as described in ref.24: Mw = 8.6x106 g/mol, Rg =

4

ACS Paragon Plus Environment

Page 5 of 20 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

52 nm and Rh =37 nm. For rod-like particles such as CNC the average length can be calculated from the average radius of gyration as L=Rg.√12. This gives an average length of 180 nm for the CNC used in this study, which is slightly smaller than the CNC used in our earlier studies. The phase diagram of mixtures of dextran and PEO was shown elsewhere

12

and did not

depend significantly on addition of NaCl or CNC at concentrations used here. Samples were prepared by mixing aqueous solutions of dextran, PEO, freshly dispersed CNC and NaCl in the required amounts using a minishaker. Subsequently, the emulsions were sonicated to fully disperse the components as described in ref.24. No effect of the order of mixing was observed for the experimental results shown here. The morphology of the emulsions in the presence of fully dispersed CNC did not depend on the time or intensity of sonication and vigorous hand shaking was sufficient. The emulsions were imaged by confocal microscopy (Leica TCS-SP2). CNC was labeled with Nile blue A by adding 10 ppm of the dye to the stock solution of 20 g/L CNC. The excess dye partitioned preferentially to the PEO phase. Rheology The solutions were loaded within a few minutes after dispersion of the CNC on a rheometer (DHR3, TA Instruments) with a cone-plate geometry (diameter 40 mm, angle 2°). Subsequently, the system was presheared during 1 min at 100 s-1 before starting the experiment. The storage (G') and loss (G") moduli were measured as a function of time (t) at 0.1 Hz in the linear response regime (1% deformation). We note that the measurement of the evolution of G' vs t at short times is unreliable when gelation is very fast, because one cannot neglect the time between ultrasonication and the start of the measurement. Systems that gel rapidly had already significantly evolved when they were loaded and the applied preshear was not sufficient to fully rejuvenate the suspensions. This was evident when we repeated the measurements after preshearing the systems that had evolved in the rheometer, which showed that in this case the systems gelled systematically more quickly. For some systems G' and G" were measured after the time sweep as a function of the frequency and as a function of the deformation at 0.1 Hz.

Results and discussion PEO in dextran emulsions containing 1.9% (w/w) PEO and 12% (w/w) dextran were prepared with different quantities of NaCl ([NaCl]= 10 - 100 mM) and CNC (C = 0.2 - 8 g/L) as described in the

5

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

Materials and Methods section. At this composition, the polymers fully phase separated with 8.2% PEO in the dispersed phase and 15.8% dextran in the continuous phase. The volume fraction of the dispersed phase was φPEO = 25%. The interfacial tension between the two phases at these conditions was found to be 75 μN/m2, see ref.4. Fig. 1 shows how emulsions containing 0.2 g/L CNC evolved visually with time. At all [NaCl] the PEO droplets creamed, but, remarkably, the rate of creaming decreased with increasing [NaCl] until 30 mM and increased when [NaCl] was increased above 40 mM. After standing overnight, the creamed PEO droplet layer had reached steady state with a relative height that increased weakly with increasing [NaCl] up till 30 mM and remained the same at larger [NaCl]. At [NaCl] = 10 and 20 mM, a fraction of the PEO droplets was unstable and formed a small continuous top layer of the PEO phase. In the absence of salt, a clear PEO top layer was formed indicating that all PEO droplets had coalesced (results not shown).

15 min

30 min

2h

12 h

Figure 1. Evolution with time of PEO in dextran emulsions containing 0.2 g/L CNC and different NaCl concentrations. From left to right [NaCl] = 100, 70, 50, 40, 30, 20, 10 mM.

Creaming was slower at C = 0.5 g/L and at C = 1 g/L it took several weeks to reach steady state. At each CNC concentration the height of the creamed layer at steady state was the same for [NaCl] > 30 mM and increased with increasing C, see Fig.2. At C ≥ 2 g/L creaming was no longer observed for [NaCl] ≥ 20 mM. However, full destabilization was still observed in the absence of salt within a few days at C= 2 and 3 g/L and partial destabilization was observed at [NaCl] = 10 mM, results not shown.

6

ACS Paragon Plus Environment

Page 6 of 20

Page 7 of 20 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

0.2 g/L

0.5 g/L

1 g/L

Figure 2. PEO in dextran emulsions containing different CNC and NaCl concentrations after having reached steady state. From left to right [NaCl] = 100, 70, 50, 40, 30, 20, 10 mM.

The microstructure of the creamed layer at steady state was studied by CLSM using fluorescently labeled CNC. Fig.3 shows CLSM images of emulsions in the presence 100 mM NaCl at different CNC concentrations between C = 0.2 g/L and 8 g/L. Similar microstructures were observed at other NaCl concentrations above 30 mM. Droplets with different sizes were formed in all systems, but as was discussed in more detail elsewhere12, the average droplet diameter decreased with increasing CNC concentration down to about 10 µm at C = 2 g/L. Adding more CNC did not have much effect on the droplet size. At C = 0.2 g/L the images show droplets covered with a layer of CNC that stick together. At C = 0.5 g/L the layer is more irregular due to small CNC aggregates that have accumulated at the interface. The CNC aggregates are much larger at C = 1 g/L and form bridges between the droplets. At higher concentrations the CNC aggregates form a network between the droplets that becomes denser with increasing concentration. We note that excess of the fluorescent probe that was used to physically label CNC preferred the PEO phase, which explains why the PEO phase is colored at low CNC concentrations.

7

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

0.2 g/L

0.5 g/L

3 g/L

5 g/L

1 g/L

Page 8 of 20

2 g/L

8 g/L

Figure 3. CLSM images (200 µm × 200 µm) of the creamed layer of PEO in dextran emulsions containing 100 mM NaCl at different CNC concentrations.

Elsewhere we showed that at all CNC concentrations a fraction of CNC covers the droplet surface and the rest is dispersed in the continuous dextran phase 12. At low C, the majority of CNC is adsorbed to the interface and there is little free CNC. In the presence of salt, free CNC binds to each other and to the CNC at the interface. Fig. 3 shows that very few aggregates are formed at C < 1 g/L, which corroborates earlier results obtained in the absence of salt. Aggregates of free CNC diffuse until they collide with the CNC at the interface and subsequently remain stuck to the surface of the droplets. The PEO droplets with their layer of aggregated CNC cream and when they collide with another droplet the CNC at the surfaces stick together forming bridges between the droplets. This process continues until the droplets together with all the CNC form a network that can resist the buoyancy of the droplets at which point creaming is arrested. When the concentration of CNC is increased, more aggregated CNC accumulates at the surface of the droplets. These aggregates form longer bridges between droplets so that the average distance between droplets increases. Therefore, the density of the network at steady state decreases with increasing CNC concentration. When the average distance between bridged droplets is sufficiently large, a space filling network of droplets crosslinked by CNC aggregates can be formed

8

ACS Paragon Plus Environment

Page 9 of 20 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

without creaming. If in addition this network is sufficiently strong to resist the buoyancy, creaming no longer takes place. Elsewhere, it was shown that the CNC network in water requires a minimal concentration of about 3 g/L to be sufficiently stiff to resist gravity

24

. At lower concentrations the CNC network

collapses unto itself until it has reached a sufficiently high density to resist gravity. Since CNC is denser than water the collapsed gel forms a layer at the bottom of the tube. In the case of the emulsions the buoyancy of the PEO droplets is stronger than the gravity of the CNC attached to the droplets. Therefore, creaming was observed instead of sedimentation. It appears from our observations that the network with the embedded PEO droplets was sufficiently strong to resist buoyancy of the droplets for C ≥ 2 g/L. However, the yield stress of the emulgels was very low, because the systems flowed when tilted even at C = 8 g/L. Even though weak gels flow under gravity they can inhibit creaming, because resistance to buoyancy requires a much smaller yield stress; less than 10-4 Pa for droplets with a radius of 10 µm and density difference of about 50 kg/m3. In this scenario we have assumed that aggregation of CNC is fast compared to creaming, which is the case at higher salt and CNC concentrations. However, at lower NaCl and CNC concentrations aggregation is slower, and creaming is faster so that we need to consider the rate of aggregation in comparison with the rate of creaming. Visual observations showed that above 30 mM NaCl the height of the creamed layer was independent of [NaCl]. This means that for [NaCl] >30 mM the CNC aggregated sufficiently fast to connect the droplets and arrest creaming at the same stage. At lower ionic strength the height decreased with decreasing [NaCl] and the droplets became increasingly unstable. In the absence of salt, the suspensions were unstable even at high C. It appears that net attractive interaction between CNC at the interface induced by adding salt was needed to avoid coalescence. Even if the final height of the creamed layer at [NaCl] > 30 mM was the same there was a clear effect of the ionic strength on the rate of creaming at C = 0.2 g/L. Creaming was faster both with increasing [NaCl] above 40 mM and with decreasing [NaCl] below 30 mM, see Fig. 1. The increasing creaming rate at [NaCl] < 30 mM can be understood by more coalescence of droplets with decreasing [NaCl] causing an increase of the droplet size during creaming. Coalescence led to macroscopic phase separation for [NaCl] < 20mM. The increasing creaming rate with increasing [NaCl] concentration above 40 mM has a different origin. Most likely, it was caused by aggregation of droplets during creaming which also increases the creaming rate, because the friction per droplet felt by aggregated droplets is less than that of individual droplets.

9

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 10 of 20

When more CNC was added, creaming became slower at all [NaCl]. The increase of the creaming rate with increasing [NaCl] above 30 mM observed at C = 0.2 g/L was not observed at C = 0.5 g/L and C = 1 g/L. At C = 0.5 g/L the creaming rate was approximately the same at [NaCl] ≥ 30 mM and at C = 1 g/L creaming was significantly slower at higher [NaCl]. One possible explanation is that when more CNC is present, the effect of aggregating droplets becomes counterbalanced by the weight and the friction of the CNC aggregates attached to the droplets.

Rheology

Gelation of CNC suspensions and emulsions was investigated by measuring the oscillatory shear moduli as a function of time at a fixed frequency of 0.1 Hz and at small deformation keeping the response in the linear regime. Fig.4 shows the evolution with time of G' and G" after dispersing 8 g/L CNC in 60 mM NaCl. G' increased strongly after a few hundred seconds and crossed G" implying that a gel was formed. After the steep increase, G' continued to increase slowly over the whole duration of the experiment indicating that bonds slowly reinforced with time. In the emulsions, the excess CNC was principally situated in the continuous phase that contained 16% dextran 12. Comparison of the evolution of the shear moduli of CNC at [NaCl] = 60 mM in the presence of or absence of 16% dextran shows that gelation was faster in dextran solutions, but that the gel stiffness was not significantly influenced by the presence of dextran, see fig. 4. It is known that dextran does not have specific interactions with CNC

25

, so slower gelation might be

expected as diffusion of CNC is slower in the semi-dilute dextran solution. We speculate that the faster gelation was caused by depletion of dextran chains, which induces an effective attraction between the CNC particles and reduces the barrier that needs to be overcome to form bonds. The higher loss modulus before gelation in the presence of 16% dextran was caused by the viscous dextran solution.

10

ACS Paragon Plus Environment

Page 11 of 20

10

G',G" (Pa)

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

water 16% dextran

1

0.1

0.01

102

103

104

t (s)

Figure 4. Evolution with time after preparation of G' (closed symbols) and G" (open symbols) at 0.1 Hz for CNC suspensions at C = 8 g/L and [NaCl] = 60 mM with and without 16% dextran.

Fig.5a shows the effect of the CNC concentration between C = 3g/L and C = 16 g/L in the presence of 16% dextran at [NaCl] = 60 mM. As expected, increasing C led to faster gelation and stiffer gels. The results at different C can be superimposed using horizontal and vertical shift factors. Fig. 5b shows the master curve of G/Gel vs t/tg where we defined tg as the time at which G' = 10-2 Pa and the elastic modulus of the gels Gel is G' at t/tg=100. For the systems studied here tg was rather short and the value was not reliable for reasons explained in the Materials and Methods section. The effect of C on tg was discussed in detail in ref. 24, where it was shown that tg decreases with increasing C following a power law: tg ∝ C-1.7. Here we focus on the effect of C on the elastic modulus. Fig. 6 shows that Gel increased steeply starting close to C = 3 g/L and more slowly at higher concentrations.

11

ACS Paragon Plus Environment

Langmuir

3 g/L 4 g/L 6 g/L 8 g/L 12 g/L 16 g/L

102

b 1

G' / Gel

101 G' (Pa)

100

0.1

10-1 a 10-2

0.01

101

102

103

104

100

101

102

t/tg

t (s)

Figure 5a. Evolution with time after preparation of G' at 0.1 Hz for CNC aqueous solutions in 16% dextran at [NaCl] = 60 mM and with different CNC concentrations indicated in the figure. Figure 5b shows the same data after normalization of G' with the elastic modulus and t with the gel time.

100

Gel (Pa)

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 20

10

1 1

2

3

4

5 6 7 8 910

20

C (g/L)

Figure 6. Elastic modulus of gels formed by aqueous suspensions of CNC at various concentrations at 60mM NaCl in 16 % dextran (circles) and emulgels at φ = 35% (squares).

The effect of varying the NaCl concentration on the gel time of CNC in the absence of dextran was already discussed in detail elsewhere24. It was shown that tg decreased with increasing NaCl

12

ACS Paragon Plus Environment

Page 13 of 20

concentration following a very steep power law dependence: tg∝ [NaCl]-10. The evolution of G' with time is plotted in fig. 7 for CNC suspensions at different [NaCl] at C = 8 g/L in the presence of 16 % dextran. As expected gelation strongly slowed down with decreasing [NaCl], but the effect on the elastic modulus was weak. This is perhaps not surprising, because the structure of the gels probed by light scattering was reported to be the same in this range of NaCl concentrations and the strength of the crosslinks between CNC is not controlled by electrostatic interactions. 102

102 a

60mM 50mM 40mM

101

b

60mM 50mM 40mM

G' (Pa)

101

G' (Pa)

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

100

100

10-1

10-1

10-2 101

102

103

100

104

101

102

t/tg (s)

t (s)

Figure 7a. Evolution with time after preparation of G' at 0.1 Hz for CNC suspensions in 16% dextran at C = 8 g/L and different NaCl concentrations indicated in the figure. Figure 7b shows the same data after normalization of G' with the elastic modulus and t with the gel time.

In first instance we studied the effect of dispersed PEO droplets on the evolution of G' with time at C = 8 g/L for PEO in dextran emulsions at different φPEO up to 35%. At higher φPEO the dispersed PEO phase started to embed small dextran droplets and phase inversion occurred at φPEO above about 45%. The compositions were chosen to lie on the same tie-line so that the interfacial tension and the concentrations of PEO and dextran in each phase were kept constant. Fig. 8a shows that G' increased more rapidly with time when the fraction of PEO droplets was higher. The results obtained at different φPEO superimposed when plotted as a function of t/tg, see fig. 8b, which means that Gel was not substantially influenced by the presence of PEO droplets at this CNC concentration. The decrease of tg with increasing φPEO is shown as an insert of fig. 8b.

13

ACS Paragon Plus Environment

Langmuir

102

b

a 10

0% 5% 10% 15% 20% 25% 35%

10-1

100 80

1

tg (s)

100

G' (Pa)

101

G' (Pa)

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 20

60 40 20 0 0

10

0.1

10-2 101

102

103

104

1

10

20

φPEO (%)

30

40

100

t/tg

t (s)

Figure 8a. Evolution with time after preparation of G' at 0.1 Hz for emulsions at C = 8 g/L and [NaCl] = 60 mM with different volume fractions of PEO droplets indicated in the figure. Figure 8b shows the same data after normalization with the gel time. The inset shows the dependence of the gel time on φPEO.

The effect of dispersed PEO droplets on gelation at different CNC concentrations was studied for φPEO = 35%, see fig. 9. Comparison with gelation in the absence of PEO, see fig.5a, confirms that gelation was systematically faster in the presence of PEO. The concentration dependence of Gel of the emulgels is compared with that of the CNC gels in 16% dextran in fig. 6. Gel was significantly larger at low CNC concentrations and could be determined even at CNC = 2 g/L whereas in the absence of PEO the gel at this concentration was too weak to probe with the rheometer.

14

ACS Paragon Plus Environment

Page 15 of 20

102

b

a 1

G' / Gel

101

G' (Pa)

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

100

2 g/L 3 g/L 4 g/L 8 g/L

10-1

0.01

10-2 101

102

103

0.1

100

104

101

102

t/tg

t (s)

Figure 9a. Evolution with time of G' at 0.1 Hz for emulsions at φPEO = 35% and [NaCl] = 60mM with different CNC concentrations indicated in the figure. Figure 9b shows the same data after normalizing G' with the elastic modulus and t with the gel time.

In the comparison we need to consider that in the emulsions the concentration of CNC in the dextran phase is different from the total concentration, because a fraction is adsorbed at the interface and the PEO phase contains almost no CNC. Elsewhere we showed that the fraction of CNC at the interface for φPEO = 25% decreased with increasing C and was about 40 % at C = 2 g/L. This means that at C = 2 g/L the concentration of CNC in the dextran phase was actually somewhat smaller than 2 g/L. The larger elastic modulus of the emulsions can therefore not be explained by an increase of the CNC concentration in the dextran phase. The increase of Gel shows that the PEO droplets embedded in the network were elastically active. Even though the interior of the droplets was liquid, they were rendered elastic by the layer of aggregated CNC that covered the droplets. With increasing CNC concentration, the elastic modulus of the CNC network in the continuous phase increased rapidly. As a consequence, the relative contribution of the droplets to the gel stiffness decreased with increasing C, see Fig. 6. The frequency dependence of all gels showed a plateau for G' at lower frequencies and a weak minimum for G" that was much smaller than G', results not shown. The measured values of G' decreased when the oscillation strain was increased above about 10% as is illustrated in fig. 10 for emulsion gels with φPEO=35% and different low CNC concentrations. The yield stress of these strong

15

ACS Paragon Plus Environment

Langmuir

shear thinning emulsion gels was at most a few Pa, which explains why they easily flowed when tilted. 102 2 g/L 3 g/L 4 g/L

101 G' (Pa)

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 16 of 20

100

10-1

10-2 1

10

100

1000

strain (%)

Figure 10. Dependence of the oscillation strain on the measured values of G' at 0.1 Hz for emulsion gels with φPEO=35% and different CNC concentrations indicated in the figure.

Conclusions Water-in-water emulsion gels can be produced by inducing aggregation of the same particles that procure Pickering-type stability against coalescence of dispersed droplets provided that the excess particle partition is predominantly to the continuous phase. By varying the concentration of the CNC, the elastic modulus of the emulsion gels could be varied continuously. In this way weak emulsion gels were produced at low CNC concentrations that inhibited creaming of the dispersed droplets, but still flowed freely under gravity. At high CNC concentrations stronger gels could be made that supported their own weight when turned upside down. When the excess of particles is small, a network is formed of droplets that are stuck to each other via the thin layers of particles that surround each droplet. When the excess is large, aggregates of particles in the continuous phase form a network in which the droplets are embedded. The droplets are elastically active and their contribution to the stiffness of the gels is important for weak gels with relatively small amounts of excess particles.

16

ACS Paragon Plus Environment

Page 17 of 20 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

The rate of aggregation of the particles could be tuned here by varying the amount of added salt. Tuning of the aggregation rate is important, because if aggregation is too slow, creaming starts before the network can be formed. If it is too fast the particles do not have enough time to cover the droplets and will instead attach to the interface in the form of aggregate without fully covering the droplet surface.

17

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

References 1. Frith, W. J., Mixed biopolymer aqueous solutions - phase behaviour and rheology. Adv. Colloid Interface Sci. 2010, 161, (1-2), 48-60. 2. Hatti-Kaul, R., Aqueous two-phase systems: methods and protocols. Springer Science & Business Media: 2000; Vol. 11. 3. Ding, P.; Wolf, B.; Frith, W. J.; Clark, A. H.; Norton, I. T.; Pacek, A. W., Interfacial tension in phase-separated gelatin/dextran aqueous mixtures. J Colloid Interface Sci 2002, 253, (2), 367-76. 4. Balakrishnan, G.; Nicolai, T.; Benyahia, L.; Durand, D., Particles trapped at the droplet interface in water-in-water emulsions. Langmuir 2012, 28, (14), 5921-6. 5. Esquena, J., Water-in-water (W/W) emulsions. Current Opinion in Colloid & Interface Science 2016, 25, 109-119. 6. Firoozmand, H.; Murray, B. S.; Dickinson, E., Interfacial Structuring in a Phase-Separating Mixed Biopolymer Solution Containing Colloidal Particles. Langmuir 2009, 25, (3), 1300-1305. 7. Firoozmand, H.; Rousseau, D., Tailoring the morphology and rheology of phase-separated biopolymer gels using microbial cells as structure modifiers. Food Hydrocolloids 2014, 42, 204-214. 8. Ganley, W. J.; Ryan, P. T.; van Duijneveldt, J. S., Stabilisation of water-in-water emulsions by montmorillonite platelets. Journal of colloid and interface science 2017, 505, 139-147. 9. Hanazawa, T.; Murray, B. S., Effect of Oil Droplets and Their Solid/Liquid Composition on the Phase Separation of Protein–Polysaccharide Mixtures. Langmuir 2013, 29, (31), 9841-9848. 10. Murray, B. S.; Phisarnchananan, N., The effect of nanoparticles on the phase separation of waxy corn starch+ locust bean gum or guar gum. Food Hydrocolloids 2014, 42, 92-99. 11. Nicolai, T.; Murray, B., Particle stabilized water in water emulsions. Food Hydrocolloids 2016, 68 157-163. 12. Peddireddy, K. R.; Nicolai, T.; Benyahia, L.; Capron, I., Stabilization of Water-in-Water Emulsions by Nanorods. ACS Macro Letters 2016, 5, 283-286. 13. Poortinga, A. T., Microcapsules from self-assembled colloidal particles using aqueous phaseseparated polymer solutions. Langmuir 2008, 24, (5), 1644-1647. 14. Vis, M.; Opdam, J.; van’t Oor, I. S.; Soligno, G.; van Roij, R.; Tromp, R. H.; Erné, B. H., Water-inWater Emulsions Stabilized by Nanoplates. ACS Macro Letters 2015, 4, 965-968. 15. Aveyard, R.; Binks, B. P.; Clint, J. H., Emulsions stabilised solely by colloidal particles. Adv. Colloid Interface Sci. 2003, 100, 503-546. 16. Binks, B. P.; Horozov, T. S., Colloidal particles at liquid interfaces. Cambridge Univ Press: Cambridge, 2006. 17. Nguyen, B. T.; Nicolai, T.; Benyahia, L., Stabilization of water-in-water emulsions by addition of protein particles. Langmuir 2013, 29, (34), 10658-10664. 18. Dickinson, E., Emulsion gels: The structuring of soft solids with protein-stabilized oil droplets. Food Hydrocolloids 2012, 28, (1), 224-241. 19. Gonzalez-Jordan, A.; Benyahia, L.; Nicolai, T., Cold gelation of water in water emulsions stabilized by protein particles. Colloids and Surfaces A: Physicochemical and Engineering Aspects 2017, 532, 332-341. 20. Habibi, Y.; Lucia, L. A.; Rojas, O. J., Cellulose Nanocrystals: Chemistry, Self-Assembly, and Applications. Chemical Reviews 2010, 110, (6), 3479-3500. 21. Klemm, D.; Kramer, F.; Moritz, S.; Lindstrom, T.; Ankerfors, M.; Gray, D.; Dorris, A., Nanocelluloses: A New Family of Nature-Based Materials. Angewandte Chemie-International Edition 2011, 50, (24), 5438-5466.

18

ACS Paragon Plus Environment

Page 18 of 20

Page 19 of 20 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

22. Araki, J., Electrostatic or steric? - preparations and characterizations of well-dispersed systems containing rod-like nanowhiskers of crystalline polysaccharides. Soft Matter 2013, 9, (16), 4125-4141. 23. Boluk, Y.; Lahiji, R.; Zhao, L.; McDermott, M. T., Suspension viscosities and shape parameter of cellulose nanocrystals (CNC). Colloids and Surfaces A: Physicochemical and Engineering Aspects 2011, 377, (1), 297-303. 24. Peddireddy, K. R.; Capron, I.; Nicolai, T.; Benyahia, L., Gelation kinetics and network structure of cellulose nanocrystals in aqueous solution. Biomacromolecules 2016, 17, (10), 3298-3304. 25. Hu, Z.; Cranston, E. D.; Ng, R.; Pelton, R., Tuning cellulose nanocrystal gelation with polysaccharides and surfactants. Langmuir 2014, 30, (10), 2684-2692.

19

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

Graphical Abstract

20

ACS Paragon Plus Environment

Page 20 of 20