Accepted Manuscript

How hydrophobically modified chitosans are stabilized by biocompatible lipid

aggregates

N. Ruocco, H. Frielinghaus, G. Vitiello, G. D’Errico, L.G. Leal, D. Richter, O.

Ortona, L. Paduano

PII: S0021-9797(15)00342-2

DOI: http://dx.doi.org/10.1016/j.jcis.2015.03.058

Reference: YJCIS 20369

To appear in: Journal of Colloid and Interface Science

Received Date: 23 January 2015

Accepted Date: 31 March 2015

Please cite this article as: N. Ruocco, H. Frielinghaus, G. Vitiello, G. D’Errico, L.G. Leal, D. Richter, O. Ortona,

L. Paduano, How hydrophobically modified chitosans are stabilized by biocompatible lipid aggregates, Journal of

Colloid and Interface Science (2015), doi: http://dx.doi.org/10.1016/j.jcis.2015.03.058

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers

we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and

review of the resulting proof before it is published in its final form. Please note that during the production process

errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

How hydrophobically modified chitosans are stabilized by

biocompatible lipid aggregates

N. Ruocco,† H. Frielinghaus,‡ G. Vitiello,ʃ, ΨG. D’Errico,¶,ΨL. G. Leal,†

D. Richter,§,‡ O. Ortona,¶,Ψ and L. Paduano*,¶,Ψ

†Department of Chemical Engineering, University of California Santa Barbara, Santa

Barbara,

California 93106-5080,

‡Jülich Centre for Neutron Science, JCNS at MLZ, Forschungszentrum Jülich GmbH, 85747

Garching, Germany,

ʃ Department of Chemical, Materials and Production Engineering, University of Naples

“Federico II”, Piazzale Tecchio 80, 80125, Naples, Italy

ΨCSGI, Consorzio interuniversitario per lo sviluppo dei Sistemi a Grande Interfase, Florence,

Italy.

¶ Department of Chemical Sciences, University of Naples “Federico II”, Complesso di

Monte S. Angelo, via Cintia, 80126, Naples, Italy.

§Jülich Centre for Neutron Science, JCNS-1, Forschungszentrum Jülich GmbH, 52425 Jülich,

Germany

*E-mail: luigi.paduano@unina.it

2

Abstract

Nanostructured hydrogels composed by biocompatible molecules are formulated and

characterized. They are based on a polymer network formed by hydrophobically modified

chitosans (HMCHIT or CnCHIT) in which vesicles of monoolein (MO) and oleic acid or

sodium oleate (NaO), depending on pH, are embedded. The best conditions for gel formation,

in terms of pH, length of the hydrophobic moieties of chitosan, and weight proportion among

the three components were estimated by visual inspection of a large number of samples.

Among all possible combinations, the system C12CHIT-MO-NaO in the weight proportion

(1:1:1) is optimal for the formation of a well-structured gel-like system, which is also

confirmed by rheological experiments. Electron paramagnetic resonance (EPR)

measurements unambiguously show the presence of lipid bilayers in this mixture, indicating

that MO-NaO vesicles are stabilized by C12CHIT even at acid pH.

A wide small angle neutron scattering investigation performed on several ternary systems of

general formula CnCHIT-MO-NaO shows that the length of the hydrophobic tail Cn is a

crucial parameter in stabilizing the polymer network in which lipid vesicles are embedded.

Structural parameters for the vesicles are determined by using a multilamellar model that

admits the possibility of displacement of the center of each shell. The number of shells tends

to be reduced by increasing the polymer content. The thickness and the distance between

consecutive lamellae are not influenced by either the polymer or MO-NaO concentration.

The hydrogel presented in this work, being fully biocompatible and nanostructured, is well-

suited for possible application in drug delivery.

3

I) Introduction

In the past and also presently, mixtures of polymers and surfactants are of major interest to

both technologists and scientists. The former are interested mainly due to the huge number of

combinations that these chemicals form and that can be used in multicomponent formulations

such as cosmetics, paints, food products, as well as drug delivery vehicles. The latter are

interested in understanding the thermodynamic reasons of their stability or instability and in

studying the microscopic structure of their aggregates [1-7]. In recent years, interest has been

mainly focused on systems formed by biocompatible polymers and surfactants [8-11]. In this

field, charged polysaccharides are of special interest due to their ready availability from

renewable sources, their biodegradability, and the absence of toxicity [12, 13]. Among these

biopolymers, chitosan plays a leading role, not only because it can be easily and plentifully

obtained from chitin deacetylation [14] but also for its antinflammatory and antimicotic

properties that are of special interest in pharmaceutical applications [15] Furthermore,

chitosan, being a charged polymer, is able to interact with charged surfactant molecules

[16,17]. Chitosan is a linear polysaccharide composed of randomly distributed D-

glucosamine and N-acetyl-D-glucosamine units. It can be obtained from crustacean shells by

deacetylation of chitin under alkaline conditions. This biopolymer can be easily modified [18,

19], due to the presence of two types of reactive groups onto which they can be grafted: the

free amine group on deacetylated units and the two hydroxyl groups on the gluco ring. The

chemical modification of chitosan can be performed in many different ways, allowing this

natural polymer to be the basis of applications in very different technological fields, such as

drug delivery, tissue engeneering, antimicrobial agents, metal ion absorption, dye removal,

viscosity control, and coating processes [20-23]. In the present report, we are concerned with

chemical modifications of chitosan. This has been obtained by grafting alkyl pendants of

different lengths to the free amine group of the saccharide unit, thus preserving its high

4

biocompatibility [20]. These chemical modifications have the aim of changing the

hydrophilic-lipophilic balance of this bio-polymer, giving rise to the formation of

hydrophobic pools that can interact with bio-compatible surfactant aggregates [24, 25]. In this

work, we investigate the interaction of these modified chitosans with anionic and non-ionic

surfactants, and analyze the microstructure of these mixtures.

Concerning the surfactants, monoolein (MO) is non-toxic, biodegradable, and biocompatible.

It has a polar head group and a non-polar hydrocarbon chain, clearly showing its amphiphilic

properties. This allows monoolein molecules to self-assemble into different liquid crystalline

structures, under varying conditions of temperature and solvent composition [26-28]. On the

basis of these properties, monoolein-based nanostructures represent a suitable strategy to

formulate useful drug delivery carriers, as well shown in literature [29-33].

Sodium oleate (NaO) is one of the main components (39% as oleic acid) of palm oil and is

the principal unsatured fat present in this natural material. As in the case of MO, NaO

aggregates in water giving rise to different structures depending on composition and

temperature [34]. In summary, both amphiphilic compounds show the tendency to aggregate

and are ecologically benign and mild. Taking into account that all compounds are

biocompatible, largely available, and affordable in the fields of food preparation and drug

delivery, we address the final goal of this paper, which is to determine the composition and

ratios among components that give rise to the formation of nanostructured gel-like systems.

In the context of their possible utilization as drug delivery systems, a quantitative analysis of

the microstructural characteristics of the CnCHIT-MO-NaO system deals with the aim to

obtain a deeper insight on the type of aggregates.

5

II) Experimental section

A. Materials

Chitosan (CHIT), classified by the supplier as medium molecular weight, batch 08028CD,

was purchased from Aldrich. Its acetylation degree was determined by NMR with a Varian

300 spectrometer to be 5%, which is in agreement with the value reported by the

manufacturer. Pentanal (97%), hexanal (98%), octanal (98%), decanal (99%), dodecanal

(92%), sodium cyanoborohydride NaCNBH3 (≥ 95%) and glacial acetic acid (~99.8%),

which were used for the alkylation procedure on the amine group, were all purchased from

Aldrich and used as received. We synthesized HMCHIT’s alkylating chitosan under mild

conditions following a reductive amination procedure suggested by Yalpani [18], and

successively modified by Rinaudo [35, 36]. The substitution degree of these alkylated

chitosans, from here on indicated as CnCHIT, is 10% as determined by nuclear magnetic

resonance (NMR), while their molecularweight is ~260kgmol-1. Monoolein (MO) and sodium

oleate (NaO) were provided by Danisco and BioChemica (Fluka), respectively, and were

used as received. For simplicity, from here on we will use the acronym NaO also in the case

of acidic solutions of sodium oleate in which only oleic acid is present. The EPR probes, 5-

doxylstearate (5-DSA) and 16-doxylstearate (16-DSA), were purchased from Sigma-Aldrich.

All solutions were prepared by weight using water obtained by inverse osmosis from an Elix

Millipore apparatus. Na2HPO4, citric acid, NaHCO3, and Na2CO3 components were used to

prepare buffer solutions with 3≤pH≤9. For SANS measurements all solutions were prepared

in heavy water (D2O) and, if necessary, in deuterated acetic acid (AcD), purchased from

Sigma-Aldrich.

B. Methods

Our aim is to study of the structures formed by CnCHIT, MO-NaO, and CnCHIT-MO-NaO

aqueous systems. The experimental approach follows three lines: i) visual inspection of

6

binary and ternary mixtures combined with a rheological analysis on the most successful

sample, ii) analysis of the structure of the aggregates by electron paramagnetic resonance

(EPR), and iii) determination of the aggregates structural parameters by small angle neutron

scattering (SANS).

1) Visual and Rheological Inspection

Several mixtures formed by CnCHIT and MO and NaO in different proportions and at

different pH conditions, were prepared by weight, left to rest overnight, and then visually

analysed.

The rheological experiments were performed with a rheometer AR-G2, which is an advanced

controlled stress instrument. The rheological flow and oscillatory experiments were made

with parallel plate geometry (diameter of 20mm). All the experiments were performed under

constant temperature, i.e. 25oC unless otherwise specified. Both frequency and flow sweeping

tests were rapidly performed after the hydrogel formation. The frequency sweeping tests were

used to obtain the viscoelastic behavior of the hydrogel in the linear regime. The experiment

was conducted isothermally in the frequency range 0.1<w<100rad/s. The stress strain was

kept constant at 1%.

The flow sweep experiments were made isothermally in the shear rate range 0.0001< <1s-1.

This range of shear rates allowed to determine the zero-shear viscosity (h0) and shear

thinning regime. The data points were collected once every 100s.

2) Electron paramagnetic resonance (EPR)

The EPR spectra were recorded using a 9 GHz Bruker Elexys E-500 spectrometer. Samples

were located in 25μL glass capillaries and flame sealed. Afterwards, the capillaries were

inserted in a quartz sample tube containing light silicone oil to provide thermal stability. The

7

experiments were carried out at 25°C. The instrumental setting for the spectra can be reported

as follows: sweep width, 120 G; resolution, 1024 points; time constant, 20.48 ms; modulation

frequency, 100 kHz; modulation amplitude, 1.0 G; incident power, 6.37 mW. The signal-to-

noise ratio was improved by collecting several scans (usually 32). The systems considered in

the present work are not paramagnetic and consequently, they are EPR silent. The EPR

analysis of the lipid aggregates microstructure was achieved by using suitable spin-probes,

such as 5-DSA and 16-DSA. The concentration of these probes in the mixtures was 5.2.10-6

mol L-1, i.e., less the 0.1 mol % of the total lipid concentration. This ratio is low enough to

avoid peak broadening due to spin exchange interaction. A semi-quantitative analysis of the

5-DSA spectra was done by measuring the outer hyperfine splitting, 2Amax, defined as the

difference (in Gauss) between the low-field maximum and the high-field minimum. 2Amax is

dependent upon both the amplitude (i.e., order) and rate of chain rotational motion, and is

therefore a useful parameter for characterizing chain dynamics in lipid aggregates [37, 38].

3) Small angle neutron scattering (SANS)

Small angle neutron scattering measurements were performed at the KWS1-FRM2-Garching

instrument with the wavelength λ=7Å and the corresponding spread Δλ/λ≤ 0.1 and at 25°C.

The two-dimensional (2D) scattering intensities in 128 by 128 channels were corrected

pixelwise for empty cell scattering, detector efficency, radial average and transformation to

absolute scattering cross sections were made with a secondary plexiglass standard [39, 40]

using QtiKWS software of the host institute. The experimental data were normalized to

absolute scale (cm-1) via calibration with the incoherent scattering [41, 42]. Two instrumental

configurations including two different collimation (C) and detecting (D) distances were used

(λ7ÅC8D8 and λ7ÅC8D2), underscripts are length in meters) to measure the scattering of the

neutrons by the samples. These configurations allowed data to be collected in the q-scattering

8

vector range: 2·10-3 Å-1<q<0.25 Å-1. In a general form, the scattering vector is defined as:

q = 4 n sin( / 2 ) , where n=1 for neutron scattering and θ is the scattering angle.

III) Results

A) Visualand Rheological Inspection

The effect of CnCHIT on the MO-NaO aqueous system was checked by visual inspection. In

Table 1, results are shown for several MO-NaO mixtures in the presence or absence of

CnCHIT’s with different lengths of the alkyl chain on the chitosan backbone. An

alphanumeric code is introduced to specify the composition of each mixture and what kind of

CnCHIT was used for each preparation. The general formula of the alphanumeric "Sample

Code" is CnCHIT(g1:g2:g3): Cn indicates the presence and the number of carbon atoms in

the chitosan alkyl chain; g1, g2, and g3 indicate the weight percentage of CnCHIT, MO, and

NaO, respectively. For example, the Sample Code C6CHIT(2:3:4) represents a ternary

mixture composed of C6CHIT at 2% w, MO at 3% w, and NaO at 4% w. HMCHIT or

CnCHIT polymers are soluble only if their amine functions are protonated, namely for pH≤ 5.

Thus, acetic acid at 1% by volume was used as the solvent for all the samples. If this is not

the case, it is indicated explicitly (see Samples 3-5 in Table 1).

Here, it is necessary to stress that, while the MO-NaO aqueous system is rather stable in

water at pH≥7, it is unstable for lower pH values (see Samples 1-5); the protonation of the

carboxyl group of the OA at low pH decreases the surface charge density of the membrane,

and also decreases the polarity of the membrane interface as widely discussed in the literature

[43]. Furthermore, the addition of hydrophobically modified chitosans to MO-NaO system

(Samples 12-28) shows different effects depending on the length of the aliphatic chain used

9

to modify CHIT. In fact, for unmodified chitosan or for short hydrophobic chains on the

polysaccharide backbone (see Samples 12-16), a precipitate is observed as a consequence of

the cited instability of MO-NaO aqueous systems at low pH values. The comparison among

samples 17 [C8CHIT(1:1:1)], 19 [C10CHIT(1:1:1)], and 23 [C12CHIT(1:1:1)] shows that by

increasing the aliphatic chain length of HMCHIT, the systems become increasingly stiff with

a gelatinous texture. In the case of Sample 23, a "wall-to-wall" gel is observed. The different

behavior of the samples: 3, 11, and 23 is shown in Fig. 1.

Figure 1: From left to right, MO-NaO (pH 9), C12CHIT and C12CHIT-MO-NaO systems(pH 4) are presented. The picture shows the increase of stiffness from (MO-NaO) or

(C12CHIT) solutions to the C12CHIT-MO-NaOgel.

This is an indication that, in spite of the instability of MO-NaO system in an acidic medium,

the presence of the CnCHIT with long aliphatic chains is able to avoid precipitation.

Furthermore, if we compare samples at increasing (MO+NaO)/C12CHIT ratios (Samples 23

and 24) or (MO+NaO)/C10CHIT ratios (Samples 19 and 20), it becomes evident that the

10

optimum weight proportion among components to get a well-structured gel is (1:1:1). The

visual analysis strongly indicates the presence of a gel-like system only at specific

compositions of CnCHIT and MO-NaO as reported in Table 1.

Therefore, the system C12CHIT(1:1:1) was selected for a detailed rheological analysis. In

Fig. 2, the oscillatory frequency experiment highlights the viscoelastic properties of the

material.

Figure 2: Dynamic oscillatory rheological characterization of C12CHIT(1:1:1). Frequency-sweep measurements provide viscoelastic insights in the linear regime

Generally, gels show G' (storage modulus) higher than G'' (loss modulus) all over the

investigated frequency range (0.1<w<100rad/s). From the mechanical spectrum, we can

distinguish between strong and weak gels [44]. A strong gel consists of moduli that are

"almost" parallel to each other and frequency independent. However, G' has a slope of zero

and G'' shows a minimum at intermediate frequencies [44, 45]. Furthermore, the storage

modulus is commonly 1-2 orders of magnitude bigger than the loss one. In soft gels, the

moduli can still be almost parallel but show a certain frequency dependence and the ratio

11

G''/G' exceeds 0.1, which is typical in soft tissues and biological gels [44, 46]. In the

C12CHIT(1:1:1) system, the averaged ratio is G''/G'~0.06 and moduli are rather low for a

well-structured gel (G’~130Pa and G’’~8Pa), thus a strict classification of the material as

weak or strong gel cannot be easily made, due to the hybrid characteristics.

In order to investigate the effect under flow of the material, a steady-state shear viscosity

analysis as function of the shear rate was performed and shown in Fig. 3.

Figure 3: Steady-state shear viscosity as function of shear rate of C12CHIT(1:1:1). Themeasurements determine the mechanical strength of the network-like system.

The shear rate dependence highlights a complex profile, which is characterized by a

Newtonian plateau, followed by intensive shear-thinning effects (~3orders of magnitude)

[47]. Here, the zero-shear viscosity is about 40600Pa.s and at intermediate shear rates the

viscosity profile shows ~ , these are characteristic values of densely organized structure.

The combination of the Newtonian plateau (low shear rates) and shear thinning effect

(intermediate and high shear rates) seems to be common to biological gels such as protein or

polysaccharide based networks [41].

12

Table 1: Visual analysis of the samples. If not expressly indicated the samples are preparedin acetic acid at 1% w.Systems prepared in water (*) and at pH 9 (**).

B) Electron paramagnetic resonance (EPR)

EPR investigation allows the morphology of the lipid aggregates (e.g., micelles vs. vesicles)

to be unambiguously identified both in aqueous solutions and in polymer hydrogels. The spin

Samplenumber

HMCHITw% MO %wt NaO

%wt Sample code Appearance

1 0.00 1.00 1.00 (0:1:1) Precipitate2 0.00 2.02 2.02 (0:2:2) Precipitate

3* 0.00 0.99 0.98 (0:1:1) Bluish fluid4** 0.00 0.52 0.50 (0:0.5:0.5) Bluish fluid5** 0.00 1.02 1.03 (0:1:1) Bluish fluid

6 1 0 0 C0 (1:0:0) Clear oil7 1.00 0.00 0.00 C5 (1:0:0) Clear oil8 1.00 0.00 0.00 C6 (1:0:0) Clear oil9 1.01 0.00 0.00 C8 (1:0:0) Clear oil

10 1 0 0 C10 (1:0:0) Clear oil11 1.01 0.00 0.00 C12 (1:0:0) Opalescent oil12 0.98 1.00 1.00 C0 (1:1:1) Precipitate13 1.01 1.00 1.01 C5 (1:1:1) Precipitate14 1.01 2.00 2.01 C5 (1:2:2) Precipitate15 0.99 1.00 1.00 C6 (1:1:1) Precipitate16 1.00 1.99 2.00 C6 (1:2:2) Precipitate17 1.00 1.00 1.00 C8 (1:1:1) Viscous oil and precipitate18 1.00 2.01 2.01 C8 (1:2:2) Viscous oil and precipitate19 1.01 1.00 1.01 C10 (1:1:1) Opaque weak gel20 1.01 2.02 2.02 C10 (1:2:2) Opaque viscous oil21 1.00 0.25 0.26 C12 (1:0.25:0.25) Opalescent viscous oil22 1.00 0.51 0.49 C12 (1:0.5:0.5) Opaque weak gel23 0.99 1.00 0.99 C12 (1:1:1) Opaque weak/strong gel24 1.02 2.01 2.00 C12 (1:2:2) Opaque weak gel25 0.13 1.02 1.04 C12 (0.1:1:1) Opaque fluid26 0.34 1.01 1.02 C12 (0.3:1:1) Opaque fluid27 0.51 1.02 1.03 C12 (0.5:1:1) Opaque viscous oil28 0.72 1.03 1.03 C12 (0.7:1:1) Opaque weak gel

13

probe 5-DSA is well-suited to this aim. 5-DSA is an amphiphilic molecule bearing a radical

nitroxide group close to the hydrophilic headgroup. Whenever inserted in supramolecular

aggregates formed by other amphiphiles, it monitors the structure and dynamics of the

aggregate/aqueous medium interphase. Particularly, 5-DSA shows an almost isotropic three-

line spectrum when embedded in micelles [48], while a well-resolved anisotropic line shape

is observed for vesicles [49].

MO-NaO (0:1:1) pH 9

C12CHIT-MO-NaO (1:1:1)

2Amax

= 50.5 ± 0.2 G

pH 4 2Amax

= 45.3 0.4 G

5-DSA

16-DSA

MO-NaO (0:1:1) pH 9

C12CHIT-MO-NaO (1:1:1) pH 4

±

2Amax

Figure 4: EPR spectra of 5-DSA and 16-DSA in the aggregates formed by MO-NaO at pH 9and C12CHIT-MO-NaO (1:1:1) at pH 4.

The spectra of 5-DSA in MO-NaO, registered both in water and in a buffer solution at pH 9,

show the anisotropic line shape typical of liposomes, see Fig. 4. No measure could be done at

acidic pH because of lipid precipitation. The presence of bilayered structure at pH 7 is in

qualitative agreement with the literature [26], in which the strong propensity of this mixture

to form a cubic phase in excess of water is shown evolving in reversed phases by increasing

14

monoolein content [50]. Nevertheless, the NaO and MO w% explored in this work is low

enough to exclude the formation of the cubic phase.

Interestingly, in the C12CHIT-MO-NaO gel, prepared at pH 4, 5-DSA also shows an

anisotropic line shape, as reported in Fig. 4. This is clear evidence that in these systems the

polymer inhibits lipid precipitation, and the bilayered structure of MO-NaO aggregates is

preserved. The significant reduction of the 2Amax value, indicated in the figure, is connected

to the different protonation state of the spin probe (i.e., protonated at pH 4 and deprotonated

at pH 9) [51] and consequently is scarcely informative on changes in bilayer fluidity.

To investigate the microstructure of lipid self-aggregates both in MO-NaO and C12CHIT-

MO-NaO systems, spectra of 16-DSA were also registered. 16-DSA bears the nitroxide group

close to the terminus of the hydrophobic tail, and thus monitors the deep interior of the lipid

aggregate. In all cases considered here, almost isotropic three-line spectra were observed,

indicating a fluid-like interior of the bilayers. A gradient in lipid ordering and dynamics in

going from the aggregate surface (monitored by 5-DSA) to its interior (monitored by 16-

DSA) is a hallmark of lipid bilayers in the liquid-crystalline phase [52, 53]. Thus, the EPR

results conclusively point to the presence of lipid bilayers, presumably in the vesicular form.

C) Small angle neutron scattering (SANS)

A SANS analysis was performed on several of the systems in Table 1 to detect the main

structural characteristics of the aggregates present in solution. SANS was used to highlight

some details at a mesoscopic level both for the binary and ternary chitosans as well as MO-

NaO aggregates.

15

1) Aggregates of hydrophobically modified chitosans

In Fig. 5, the scattering profiles of four hydrophobically modified chitosans are reported. A

quantitative analysis of the SANS data excludes the presence of compact micellar aggregates,

as for classical surfactants, thus indicating the presence of a more complex aggregation

system. The aggregation ability of these polymers has been highlighted in the past by

fluorescence studies [25]. According to the analysis of Esquenet [54] based on the work of

Semenov [55], these systems were interpreted as a network of intermolecularly bridged

flower-like aggregates.

10-4

10-2

100

102

104

106

0.001 0.01 0.1

x100

x101

x102

x103q-2.6

q-2.6

q-2.7

q-3.6

q-0.3»q-0.9:

q-1.2:

q-3.6

····

C12CHIT (1:0:0)C8CHIT (1:0:0)C6CHIT (1:0:0)C5CHIT (1:0:0)

(d/d

cm-1

1-q/Å

Figure 5: Scattering cross sections of hydrophobically modified chitosans with differentlength of the aliphatic chains at pH 4 and 25°C. Cross sections have been multiplied with ascale factor to allow a direct comparison. Error bars result inside the plotted points

On the basis of this interpretation and due to the lack of a quantitative theory, the discussion

will be limited to some qualitative aspects. Owing to the lack of very low q-region (the

Guinier region) for all samples, we can only state that the size of the network is above

experimental 2 /qmin, which is of the order of hundreds of Å. In the low q-range, the power

16

law of the scattering profile evolves from a ~q-2.6for the short chained polymers to a ~q-3.6in

the case of C12CHIT, suggesting a transition of the aggregate morphology.

The latter law corresponds to the presence of aggregates that are more compact. Furthermore,

in the intermediate q-range, 0.01Å-1<q<0.04 Å-1, the variation of the scattered intensity with q

shows a transition from a ~q-1observed for C6CHIT to a ~q-3.6dependence in the case of the

more hydrophobic C12CHIT. It may be recalled that a q-1dependence corresponds to the

characteristic rod-like behavior of the semi-rigid chitosan chains, whenever we look at length

scales that are smaller than the persistence length [56]. The crossover, at which the transition

between a ~q-3and ~q-1regime is observed, provides a qualitative evaluation of the length

scale at which the polymer begins to fold. Interestingly, the q-1 behavior is present only in the

case of chitosan polymers with a hydrophobic pendant of intermediate length, namely C6 and

C8CHIT. The absence of the q-1power law for the polymer with the longest pendant,

C12CHIT, is an indication that a more compact network dominates over a rod-like scattering

profile. In contrast with this observation, in the case of the polymer with the shorter aliphatic

chain (C5CHIT), this behavior is probably due to the high flexibility of the backbone in the

absence of well-structured hydrophobic pools that are able to stiff the chain.

2) Aggregates of Monoolein-Sodium oleate in solution

In Fig. 6, the binary system MO-NaO (0:1:1) is represented at two different pH conditions.

It can be observed that the different pH values deeply modify the scattering profiles. At pH 7

and low q-values, the Guinier region seems to be detected. In the case of MO-NaO system at

pH 9, the Guinier region is not observable, leading to the conclusion that we are in the

presence of large aggregates with complex morphology.

17

10-2

10-1

100

101

102

103

104

0.01 0.11-q/Å

(d/d

cm-1

x100

x101

q-3

q-0.7MO-NaO (0:1:1) pH 7

MO-NaO (0:1:1) pH 9·

:

:

Lamellarinterference

Figure 6: Scattering cross section of MO-NaO aggregates at pH 7 and 9. Solid linescorrespond to the values of the fitting model applied. Error bars result inside the plottedpoints

The scattering profile decay suggests that unilamellar vesicles dominate at pH 7, although

some multilamellar vesiscles (with a small number of layers) may be present . Whereas, at pH

9, the system is dominated by large and multilamellar vesicles as suggested by the power law,

~q-3, and confirmed by the presence of a lamellar interference peak at higher q-values [57].

On the basis of such evidence, the system at pH 7 was modeled as a collection of unilamellar

vesicles [58] while at pH 9 a multilamellar model [59] was preferred. The fitting parameters

are collected in Table 2. As a general comment we specify that in no one of the data analysis

by theoretical models of this paper we took into account the instrumental resolution.

In the case of the unilamellar vesicle model, a dilute system is considered, in which the

structure factor S(q) is 1 and the form factor P(q) provides the main contribution to the

scattering cross section. P(q) for a sphere with one shell can be written as:

18

( ) ( ) ( ) ( )1 1 2 1 1 2 2 1 2

1 2

23 - 3 -

( ) solv

shell

V J qR V J qRkP qV qR qR

é ù= +ê ú

ë û (1)

where k is a scale factor, Vshell is the volume of the shell, V1and V2 are the volume of the core

and the total volume, respectively. R1and R2 are the radii of the core and of the whole vesicle.

The values of ρsolv, ρ1, and ρ2are the scattering length densities of the solvent, of the shell and

of the core, respectively. 1( )J x is the Bessel spherical function of the 1storder:

( ) 2

sin - cosi

x x xJ xx

= (2)

The modeling approach assumes a Schulz-Zimm distribution function and a polydisperse

inner radius, while the thickness of the bilayer has negligible polydispersity in comparison

with R2. On the other hand, the oligolamellar system at pH 9 was interpreted by using the

Kotlarchyk-Ritzau model for the scattering of lamellar stacks with a paracrystalline distortion

[59]. For this model the form factor is represented by:

( ) ( )( )

22 2

22

sin / 21( ) 2/ 2

qP q S

q q= (3)

In Eq. 3, is the scattering length density difference between sheet and the solvent, S is the

area of the sheet per unit sample volume and the bilayer thickness. According to this

model, the structure factor is also able to extract parameters such as: the number of layers, N,

and their mean separation, l, as explained in references [59, 60]. As noted above, the low q-

regime in the Guinier region is missing, so the overall size of the vesicles cannot be obtained.

However, the scattering profile provides crucial information on the bilayer length scale

(middle q-region). The model for the pH 9 system considers that the thickness of the bilayer

and the distance l between consecutive lamellae can be polydisperse. The thickness is a

Gaussian function, which is described by the average value < > and the standard

19

deviation, , while l is a Schulz distribution characterized by the average value <l>. It can

be noted from the thickness of the bilayer that the amphiphiles are interdigitated, i.e. the

aliphatic chains are partially interpenetrated. In fact, the length of the 17 carbon hydrophobic

chain of the pendant is roughly 22Å, so that in the absence of interdigitation the bilayer

thickness would be at least 44Å, i.e. larger that the 30±2Å fitting value. This interdigitation is

also confirmed by the data on bilayer monoolein thickness in a cubic lattice [61] present in

literature.

Code pH R2/[Å] < > /[Å] N l< > /[Å]

(0:1:1) 7 290±20 28±2 1 -

(0:1:1) 9 N/A 30±2 2 135±3

Table 2: Structural parameters for MO-NaO aggregatesat pH 7and 9 at 25 °C.

3) Aggregates of hydrophobically modified chitosans and monoolein-sodium oleate in

solution

In Fig. 7, the profiles of the systems CnCHIT-NaO-MO (1:1:1) in AcD with different lengths

of the aliphatic chains are reported. It must be stressed here that in the absence of any HM-

chitosan and in acidic conditions, the system containing only MO and NaO is instable due to

oleate protonation. If HMCHIT is present, partial precipitation is still observed for systems

with the shorter aliphatic tails, namely C6-MO-NaO and C8-MO-NaO while a stiff gel is

obtained with C10-MO-NaO and C12-MO-NaO, confirming once more that the MO-NaO

aggregates are stabilized at low pH by the presence of only long-tailed HM-chitosan

molecules. The scattering data of C6 and C8CHIT ternary systems refer to the acidic

20

surnatant solutions. Nevertheless, all systems are strongly characterized by the presence of a

broad correlation peak at a high q-values.

10-2

100

102

104

106

108

10-3 10-2 10-11-q/Å

(d/d

cm-1

x100

x101

x102

x103

· C12CHIT-MO-NaO (1:1:1)

···

C10CHIT-MO-NaO (1:1:1)C8CHIT-MO-NaO (1:1:1)C6CHIT-MO-NaO (1:1:1)

q-2.94

q-2.74

Figure 7: Scattering cross section for CnCHIT-NaO-MO (1:1:1) aggregates for differentlengths of the aliphatic chain at pH 4 and 25 °C. Solid lines correspond to the values of theapplied fitting model. The systems C6CHIT-NaO-MO and C8CHIT-NaO-MO show aprecipitate, i.e. phase separation. Error bars result inside the plotted points

It will be noted that these SANS results are very different from the systems shown in Fig. 6,

which only contain oleine and sodium oleate at pH 7 and 9, respectively. Those were

interpreted just using a simple unilamellar vesicle model at pH 7 and a concentric multi-

lamellar model for the system at pH 9. In the case of the ternary system, we require a more

complex multilamellar model, proposed in the past [61, 62]. We will refer to this as the

Frielinghaus model. It is the only model that, to the best of our knowledge, can give an

explanation for the presence of the correlation peak at high q-values. This is attributed to the

fact that the center of each shell of the vesicle is allowed to shift with respect to its

21

predecessor. In this way, the hydrophobic pendants of the chitosan backbone in the ternary

systems may have full access to the bilayer of the MO-NaO vesicular shells, producing a

displacement of their center. The longer and more hydrophobic the pendant of the chitosan,

the more efficent its access to the bilayer and the stronger is this displacement. Due to the

complexity of the Frielinghaus model, we refer to the original paper for a more detailed

discussion [63]. In Fig. 7, the solid lines represent the results of this model, which is valid

only in the case of the more hydrophobic modified chitosans. In fact, short aliphatic pendants

allow precipitation.

The main parameters obtained by application of the Frielinghaus multilamellar model are

reported in Table 3, in which is the solutes volume fraction.

Code N < > /Å l< > /Å

C10CHIT(1:1:1) 0.0245 9±1 21±1 39±6

C12CHIT(1:1:1) 0.0269 7±1 24±1 39±6

Table 3: Main parameters for C10CHIT-NaO-MO and C12CHIT-NaO-MO aggregates atpH 4 and 25 °C.

In Fig. 8, the scattering data for C12CHIT (1:1:1) is compared to that of a C12CHIT(1:0:0)

aqueous solution. This comparison highlights that the qualitative “flower micelle” model

cannot be used to give an interpretation of the experimental results for the ternary system,

mainly due to the presence of the broad correlation peak at q~0.16 Å-1.

22

10-3

10-1

101

103

105

10-3 10-2 10-1

(d/d

cm-1

· C12CHIT-MO-NaO (1:1:1)C12CHIT (1:0:0)

q/A-1

Figure 8: Comparison between the scattering cross section of C12CHIT (1:0:0) andC12CHIT-NaO-MO (1:1:1) aggregates at pH 4 and 25 °C. Error bars result inside theplotted points

The characterization of the ternary systems was carried out also analysing scattering data at

fixed C12CHIT composition, but different MO and NaO contents. In Fig. 9, the SANS

profiles are reported. It is shown that the scattering maximum at q~0.16 Å-1 increases with

increasing (MO+NaO)/C12CHIT ratio.

By tuning the (MO-NaO)/C12CHIT ratio, it can be noted that the q-dependency in the

intermediate region, 5.5·10-3 Å-1<q<2.5·10-2 Å-1 for all the systems is almost constant, even

slightly decreasing for low MO-NaO content. This indicates a domain fully dominated by the

contribution of the C12CHIT network.

23

10-3

10-1

101

103

105

107

10-3 10-2 10-1

(d/d

cm-1

qMax

q-3.0

q-3.0

q-3.3

q-3.6

q-3.6

1-q/Å

x100

x101

x102

x103

x100

·····

C12CHIT-MO-NaO (1:2:2)C12CHIT-MO-NaO (1:1:1)C12CHIT-MO-NaO (1:0.5:0.5)C12CHIT-MO-NaO (1:0.25:0.25)C12CHIT (1:0:0)

Figure 9: Scattering cross section for C12CHIT-NaO-MO (1:g2:g3) aggregates at pH 4and 25 °C. Solid lines correspond to the values of the applied fitting model. Error barsresult inside the plotted points

In the case of more diluted vesicles, the peak at high q-values is less pronounced, due to a

low amount of surfactants. Both C12CHIT(1:0.25:0.25) and C12CHIT(1:0.5:0.5) provide

similar values for the number of shells, whereas this value decreases with increasing

surfactants concentration, e.g. the system C12CHIT(1:2:2) results in almost unilamellar

vesicles. This set of experiments shows the importance of the surfactants and C12CHIT

concentrationon the overall stability of the systems.

On the other hand, the fitted distance between bilayers assumes comparable values for all of

them, and it is in line with the values of max2 / q , which provides the periodicity of the

scattering planes. The main structural parameters obtained by the best fit of the applied model

[63] are reported in Table 4.

24

Code N < > /Å l< > /Å

C12CHIT(1:0.25:0.25) 0.00461 10±1 21±1 39±6

C12CHIT(1:0.5:0.5) 0.0106 13±1 20±1 39±6

C12CHIT(1:1:1) 0.0269 7±1 24±1 39±6

C12CHIT(1:2:2) 0.0522 2±1 33±1 41±6

Table 4: Main structural parameters measured for C12CHIT-NaO-MO (1:g2:g3) aggregatesat pH 4 and 25 °C

The whole of these experimental results can be reasonably interpreted to suggest the

existence of two zones for the C12CHIT(1:0.25:0.25) and C12CHIT(1:0.5:0.5) systems: in

the first one the polymer HMCHIT interacts directly with the vesicles, determining a

backbone-bilayer connection and linking consecutive bilayers in the single multilamellar

aggregate; in the secondzone the polymer does not have any connection with the vesicles. By

increasing the MO-NaO contentas for the C12CHIT(1:1:1) and C12CHIT(1:2:2) systems, the

two zones disappear, converging to a single zone whereas vesicles with a lower number of

shells are present. The reason for this behavior could be ascribed to the greater surfactants

concentration, which is large enough to occupy all HMCHIT locations. Thus, the systems

have the tendency to form vesicles with a minor number of shells. Another set of

measurements, which helps to fully understand the matter is reported in Fig. 10, wherein the

system C12CHIT-MO-NaO in AcD for various polymer concentrations (from 0.1%w to

1%w) and fixed MO-NaO composition, were investigated. The results of fitting this data,

shown in Table 5, indicate that the number of shells tends to decrease if the polymer

concentration is is increased. Nevetheless, thickness and inter-distance separation of the

bilayers are comparable with the previous experiments.

25

10-2

100

102

104

106

108

10-3 10-2 10-11-q/Å

(d/d

cm-1

C12CHIT-MO-NaO (1:1:1)C12CHIT-MO-NaO (0.7:1:1)C12CHIT-MO-NaO (0.5:1:1)C12CHIT-MO-NaO (0.3:1:1)C12CHIT-MO-NaO (0.1:1:1)

···

··

x100

x101

x102

x103

x104

Figure 10: Scattering cross section for C12CHIT-NaO-MO (g1:1:1) aggregates at pH 4 and25 °C. Solid lines correspond to the values of the applied fitting model. Error bars resultinside the plotted points

Code N < > /Å l< > /Å

C12CHIT(0.1:1:1) 0.0106 15±2 20±1 37±6

C12CHIT(0.3:1:1) 0.0125 13±1 20±1 38±6

C12CHIT(0.5:1:1) 0.0130 16±2 20±1 38±6

C12CHIT(0.7:1:1) 0.0129 9±1 21±1 39±6

C12CHIT(1:1:1) 0.0269 7±1 34±1 39±6

Table 5: Main parameters of C12CHIT-NaO-MO (g1:1:1) aggregates at pH 4 and 25 °C.

26

The ternary system C12CHIT-MO-NaO is strongly influenced by the polymer concentration.

For low C12CHIT concentrations, the surfactant aggregates are separated from the chitosan

backbone. The reason for this behavior is because the amount of polymer is not high enough

to build up locations able to contain the vesicles. On the other hand, at higher concentration,

the C12CHIT can form a polymeric network, which incorporates the vesicles. This means

that, by tuning the polymer content together with the amount of MO-NaO, we have direct

control of the system, e.g. number of lamellae N and vesicular dimensions.

Conclusions

Finally, we conclude that the visual and rheological analysis allowed to determine the best

conditions for gel formation by varying pH, length of the hydrophobic chain of the CnCHIT,

and the weight ratio between the three components. The optimum condition was obtained for

the system C12CHIT-MO-NaO in acetic acid 1%v, with 1% by weight for each component

(1:1:1). The use of C12CHIT improves the conditions for aggregation of the system, due to

its long hydrophobic pendant; the MO-NaO weight ratio equal to 1 insures the presence of

vesicular aggregates, and the acidic pH provides the conditions necessary to solubilize the

polymer and build up a network-like system. Overall, it appears that by increasing the

aliphatic chain length the system becomes more rigid, adopting a gel texture for C12CHIT.

As regards the MO-NaO content, there is a precise (MO+NaO)/ C12CHIT weight ratio, i.e. 2,

that insures the formation of a gel. In fact, for both higher and lower ratios, namely 0.5 and 4,

the gels become less rigid and stiff. We can speculate that for (MO+NaO)/ C12CHIT< 2 the

number of vesicles is too low with respect to the number of the Cn chains on the polymer

backbone. Therefore, the formation of a three dimensional network due to the insertion of the

aliphatic chains inside the MO-NaO bilayer is less efficient. In contrast, for (MO+NaO)/

C12CHIT>2, the number of vesicles is too high in comparison to the aliphatic chain

27

concentration, so that a large number of vesicles cannot take part to the three dimensional

network.

The frequency and flow sweeping tests have showed the main macrostructural and

mechanical properties of the sample C12CHIT-MO-NaO (1:1:1), resulting in a material with

polyvalent characteristics. Indeed, the sample provided hybrid viscoelastic properties

(soft/strong gel) and common features of well-structured biological systems (Newtonian

plateau and shear thinning).

The EPR results unambiguously show the presence of bilayers in the C12CHIT-MO-NaO

(1:1:1) system, highlighting the stability of these structures even in acidic pH conditions. A

very complete SANS analysis was performed on several ternary systems of the general

formula CnCHIT-MO-NaO. The results highlighted and confirmed that the length of the

hydrophobic tail Cn is a crucial parameter to stabilize the polymer network, probably

organized as a system of flower-like aggregates. In the absence of chitosans the system MO-

NaO is stable only in non-acidic conditions and is organized as bilayer or vesicular structures.

The presence of long tailed chitosans allows the stabilization of the MO-NaO aggregates

even in acidic conditions. Indeed, the tuning of the components in the ternary system leads to

the conclusion that among all possible combinations, the system C12CHIT-MO-NaO in the

weight proportion 1:1:1 is optimal for the formation of a very stiff gel, in which multilamellar

vesicles of MO-NaO are embedded.

For ternary systems, structural parameters for the vesicles were determined using a

multilamellar model that admits the possibility of displacement of the center of each shell.

The thickness and the distance between consecutive lamellae are not influenced by either the

polymer concentration or the MO-NaO composition, while the number of shells and

consequently the vesicle dimensions increase by decreasing the surfactant content at constant

C12CHIT w/w% as well as by decreasing (MO+NaO) w/w% at constant chitosan

28

concentration. A possible explanation could be that the high surfactant content leads to

smaller compartments. This offers more options for polymer anchoring, while high polymer

content tends to smaller compartments too, through stronger binding of surfactants.

Ultimately, the dimensions and the number of vesicle shells can be modulated at will by

varying surfactants and chitosan composition, thus creating a flexible system very suited for

delivery of hydrophobic and hydrophilic drugs.

29

Bibliography

[1] Roscigno, P.; Paduano, L.; D’Errico, G.; Vitagliano, V. Langmuir, 2001, 17, 4510-4518.

[2] Lee, J. H.; Gustin, J. P.; Chen, T.; Payne, G. F.; Raghavan, S. R. Langmuir, 2005, 21, 26-

33.

[3] Chen, Y.; Javvaji, V.; MacIntire, I. C.; Raghavan, S. R. Langmuir, 2013, 29, 15302-

15308.

[4] Huang, W.; Wang, Y.; Zhang, S.; Huang, L.; Hua, D.; Zhu, X. Macromolecules, 2013, 46,

814−818.

[5] Lott, J. R.; McAllister, J. W.; Arvidson, S. A.; Bates, F. S.; Lodge T. P.

Biomacromolecules, 2013, 14, 2484−2488.

[6] Dowling, M. B.; Kumar, R.; Keibler, M. A.; Hess, J. R.; Bochicchio, G. V; Raghavan, S.

R. Biomaterials, 2011, 32, 3351-3357.

[7] Hao, X.; Liu, H.; Xie, Y.; Fang C.; Yang, H. Colloid Polym.Sci., 2013, 291, 1749-1758.

[8] Bao, H.; Li, L.; Gan, L. H.; Zhang, H. Macromolecules, 2008, 41, 9406-9412

[9] Mangiapia, G.; Ricciardi, R.; Auriemma, F.; De Rosa, C.; Lo Celso, F.; Triolo, R.;

Heenan, R. K.; Radulescu, A.; Tedeschi, A. M.; D'Errico, G.; Paduano, L. J. Phys. Chem. B,

2007, 111, 2166–2173.

[10] Lee, J. H.; Agarwal, V.; Bose, A.; Payne, G. F.; Raghavan, S. R. Phys. Rev. Lett., 2006,

96, 048102.

[11] Chiappisi, L.; Prevost, S.; Grillo, I.; Gradzielski, M. Langmuir, 2014, 30, 10608-10616.

[12] Chiappisi, L.; Prevost, S.; Grillo, I.; Gradzielski, M. Langmuir, 2014, 30, 1778-1787.

[13] Alonso-Sande, M.; Cuña, M.; Remuñán-López, C.; Teijeiro-Osorio, D.; Alonso-Lebrero,

J. L.; Alonso, M. J. Macromolecules, 2006, 39, 4152-4158

[14] Tolaimate, A.; Desbrieres, J.; Rhazi ,M.; Alagui ,A.; Vincendon ,M.; Vottero, P.

Polymer, 2000, 41, 2463–2469.

[15] Liu, S.; Maheshwari, R.; Kiick, K. L. Macromolecules, 2009, 42, 3-13.

[16] Zhu, C.; Lee, J. H.; Srinivasa, R.; Raghavan, S. R.; Payne, G. F. Langmuir, 2006, 22,

2951-2955.

[17] Mangiapia, G.; Frielinghaus, H.; D’Errico,G.; Ortona, O.; Sartorio, R.; Paduano, L.

Phys. Chem. Chem. Phys., 2007, 9, 6150–6158.

[18] Yalpani, M.; Hall, L.D. Macromolecules, 1984, 17, 272-281.

30

[19] Marcasuzaa, P.; Reynaud, S.; Ehrenfeld, F.; Khoukh, A.; Desbrieres, J.

Biomacromolecules, 2010, 11, 1684–1691.

[20] Lim, S-H.; Hudson, S.H. J. Macromol. Sci., Part C, Polym. Rev., 2003, 43, 223-269

[21] Ge, Y.; Zhang, Y.; He, S.; Nie, F.; Teng, G.; Gu, N. Nanoscale Res. Lett., 2009, 4, 287-

295.

[22] Rassaei, L.; Sillanpaa, M.; Edler, K. J.; Marken, F. Electroanalysis, 2009, 21, 261-266.

[23] Zemljič, L. F.; Peršin, Z.; Stenius, P. Biomacromolecules, 2009, 10, 1181–1187.

[24] Mangiapia, G.; Frielinghaus, H.; D’Errico, G.; Ortona, O.; Sartorio, R.; Paduano, L.;

Phys. Chem. Chem. Phys., 2007, 9, 6150–6158.

[25] Ortona, O.; D’Errico, G.; Mangiapia, G.; Ciccarelli, D. Carbohydr. Polym., 2008, 74,

16-22.

[26] Borné, J.; Nylander, T.; Khan, A. Langmuir, 2001, 17, 7742-7751.

[27] Ferreira, D. A.; Bentley, M. V. L. B.; Karlsson, G.; Edwards, K. Int. J. Pharm., 2006,

310, 203-212.

[28] Angelova, A.; Angelov, B; Mutafchieva, R.; Lesieur, S.; Couvreur, P. Acc. Chem. Res.,

2011, 44, 147-156.

[29] Angelov, B.; Angelova, A.; Filippov, S.; Drechsler, M.; Stepanek, P.; Lesieur, S. ACS

Nano, 2014, 8, 5216-5226.

[30] Angelov, B.; Angelova, A.; Garamus, V. M.; Drechsler, M.; Willumeit, R.; Mutafchieva,

R.; Stepanek, P.; Lesieur, S., Langmuir, 2012, 28, 16647-16655.

[31] Angelov, B.; Angelova, A.; Papahadjopoulos-Sternberg, B.; Hoffmann, S. V.; Nicolas,

V.; Lesieur, S. J. Phys. Chem. B, 2012, 116, 7676-7686.

[32] Angelov, B.; Angelova, A.; Filippov, S.; Karlsson, G.; Terrill, N.; Lesieur, S.; Stepanek,

P. Soft Matter, 2011, 7, 9714-9720.

[33] Angelov, B.; Angelova, A.; Vainio, U.; Garamus, V. M.; Lesieur, S.; Willumeit, R.;

Couvreur P. Langmuir, 2009, 25, 3734-3742.

[34] Antunes, F.E.; Coppola, L.; Gaudio, D.; Nicotera, I.; Oliviero, C. Colloids and Surfaces

A: Physicochem. Eng. Aspects, 2007, 297, 95–104.

[35] Rinaudo, M.; Auzely, R.; Vallin, C.; Mullagaliev, I. Biomacromolecules, 2005, 6, 2396-

2407.

[36] Desbrieres, J.; Martinez, C.; Rinaudo, M. Int. J. Biol. Macromol., 1996, 19, 21-28.

[37] Lange, A.; Marsh, D.; Wassmer, K. H.; Meier, P.; Kothe, G. Biochemistry, 1985, 24,

4383–4392.

31

[38] Vitiello, G.; Fragneto, G.; Petruk, A. A.; Falanga, A.; Galdiero, S.; D’Ursi, A. M.;

Merlino, A.; D’Errico G. Soft Matter, 2013, 9, 6442–6456.

[39] Russell, T. P.; Lin, J. S.; Spooner, S.; Wignall, G. D. J. Appl. Cryst., 1988, 21, 629-638.

[40] Wignall, G. D.; Bates, F. S. J. Appl. Cryst., 1987, 20, 28-40.

[41] Ruocco, N.; Dahbi, L.; Driva, P.; Hadjichristidis, N.; Allgaier, J. ; Radulescu, A.; Sharp,

M;. Lindner, P.; Straube, E.; Pyckhout-Hintzen, W.; Richter, D. Macromolecules, 2013, 46,

9122-9133.

[42] Frielinghaus, H.; Frielinghaus, X.; Ruocco, N.; Allgaier, J.; Pyckhout-Hintzen, W.;

Richter, D. Soft Matter, 2013, 9, 10484-10492.

[43] Aota-Nakano, Y; Li, S.J.; Yamazaki, M. Biochimica et Biophysica Acta , 1999, 1461,

96-102.

[44] Clark A. H.; Murphy, S. B. Adv. Polym. Sci., 1987, 83, 57-192.

[45] Hua, X.; Tang, Y.; Wang, Q.; Li, Y.; Yang, J.; Du, Y.; Kennedy J. F. Carbohydr.

Polym., 2011, 83, 1128–1133.

[46] Barbucci, R. “Hydrogels: Biological Properties and Applications”, Ed. Springer, 2010,

Milan.

[47] Popescu, M. T.; Mourtas, S.; Pampalakis, G.; Antimisiaris, S. G.; Tsitsilianis; C.

Biomacromolecules, 2011 12, 3023-3030.

[48] Kang, Y. S.; Kevan, L. J. Phys. Chem.,1994, 98, 7624–7627.

[49] McCarney, E. R.; Armstrong, B. D.; Kausik, R.; Han, S. Langmuir, 2008, 24, 10062-

10072.

[50] Angelov, B.; Angelova, A.; Mutafchieva, R.; Lesieur, S.; Vainio, U.; Garamus, V. M.;

Jensena, G. V.; Pedersena, J. S. Phys. Chem. Chem. Phys., 2011, 13, 3073–3081.

[51] Sulkowski, W. W.; Pentak, D.; Nowak, K.; Sulkowska, A. Journal of Molecular

Structure, 2006, 792-793, 257-264.

[52] Vitiello, G.; Falanga, A.; Galdiero, M.; Marsh, D.; Galdiero, S.; D'Errico, G. Biochim.

Biophys. Acta Biomembranes,2011, 1808, 2517–2526.

[53] Vitiello, G.; Di Marino, S.; D’Ursi, A. M.; D'Errico, G. Langmuir, 2013, 29, 14239–

14245.

[54] Esquenet, C.; Terech, P.; Boué F.; Buhler, E. Langmuir, 2004, 20, 3583-3592.

[55] Semenov, A. N.; Rubinstein, M. Macromolecules, 2002, 35, 4821-4837.

[56] Muller, F.; Manet, S.; Jean, B.; Chambat, G.; Boué, F.; Heux, L.; Cousin, F.

Biomacromolecules, 2011, 12, 3330–3336.

32

[57] Mangiapia, G.; D’Errico, G.; Simeone, L.; Irace, C.; Radulescu, A.; Di Pascale, A.;

Colonna, A.; Montesarchio, D.; Paduano, L. Biomaterials, 2012, 33, 3770-3782.

[58] Guinier, A.; Fournet, G. "Small-Angle Scattering of X-Rays", John Wiley and Sons,

1955, New York.

[59] Kotlarchyk, M.; Ritzau, S. M. J. Appl. Cryst.,1991, 24, 753-758.

[60] Harvey, R. D.; Barlow, D. J.; Drake,A. F.; Kudsiova, L.; Lawrence, M. J.; Brain, A. P.

R.; Heenan, R. K. J. Colloid Interf. Sci., 2007, 315, 648-661.

[61] Angelov, B; Angelova, A.; Garamus, V. M.; Lebas, G.; Lesieur, S.; Ollivon, M; Funari,

S. S.; Willumeit, R.; Couvreur, P. J. Am. Chem. Soc., 2007, 129, 13474-13479.

[62] Angelova, A.; Angelov, B.; Garamus, V. M.; Couvreur, P.; Lesieur, S. J. Phys. Chem.

Lett., 2012, 3, 445-457.

[63] Frielinghaus, H. Phys. Rev. E, 2007, 76, 0516031-0516038.

33

Graphical