Cationic nanofibrillar cellulose with high antibacterial properties

Cationic nanofibrillar cellulose with high antibacterial properties

Accepted Manuscript Title: Cationic Nanofibrillar Cellulose with High Antibacterial Properties Author: Achraf Chaker Sami Boufi PII: DOI: Reference: ...

2MB Sizes 0 Downloads 51 Views

Accepted Manuscript Title: Cationic Nanofibrillar Cellulose with High Antibacterial Properties Author: Achraf Chaker Sami Boufi PII: DOI: Reference:

S0144-8617(15)00490-7 http://dx.doi.org/doi:10.1016/j.carbpol.2015.06.003 CARP 9981

To appear in: Received date: Revised date: Accepted date:

21-4-2015 26-5-2015 1-6-2015

Please cite this article as: Chaker, A., and Boufi, S.,Cationic Nanofibrillar Cellulose with High Antibacterial Properties, Carbohydrate Polymers (2015), http://dx.doi.org/10.1016/j.carbpol.2015.06.003 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 2 3 4

►Cationic

5

homogenization► the quaternization of the pulp was shown to facilitate the defibrillation

6

processes► the cationic NFC was composed of individual fibrils 4-5 nm in width and length

7

in the micronic scale was ► the inclusion of C-NFC was shown to efficiently enhance the

8

antibacterial activity

Research highlights (C-NFC)

cellulose

was

prepared

via

a

high

pressure

ip t

nanofibrillar

Ac ce p

te

d

M

an

us

cr

9

1 Page 1 of 29

9

Cationic Nanofibrillar Cellulose with High Antibacterial Properties

10 Achraf Chakera, Sami Boufia*

11 a

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

ip t

12 13

* Corresponding author: Sami Boufi University of Sfax-Faculty of science BP 1171-3000 Sfax-Tunisia Fax: 216 74 274 437 E-mail [email protected]

21

Abstract

22

Cationic nanofibrillar cellulose (C-NFC) was prepared via a high pressure homogenization

23

using quaternized cellulose fibers with glycidyltrimethylammonium chloride. It has been

24

shown that the quaternization of dried softwood pulp facilitated the defibrillation processes.

25

The effects of the trimethylammonium chloride content on the fibrillation yield, the

26

transparency degree of the gel, the rheological behaviour of the NFC suspension and their

27

electrokinetic properties were investigated. AFM observation showed that the NFC

28

suspension consisted of individualized cellulose I nanofibrils 4-5 nm in width and length in

29

the micronic scale. In addition to their strong reinforcing potential, the inclusion of C-NFC

30

into a polymer matrix was shown to efficiently enhance the antibacterial activity. The

31

reinforcing potential of C-NFC, studied by dynamic mechanical analysis (DMA), was

32

compared to anionic NFC and the difference was explained in terms of the nanofibrils

33

capacities to build up strong networks held by hydrogen bonding.

34

Key words: nanofibrillar cellulose, nanocomposites, cationic nanofibrils, antibacterial.

Ac ce p

te

d

M

an

us

cr

14 15 16 17 18 19 20

2 Page 2 of 29

1. Introduction

36

The production of nanocellulose from natural cellulose resources is undoubtedly one of the

37

most breakthrough in cellulose based materials over the last two decades (Khaliln Bhat,&

38

Yusra 2012, Kalia et al 2014). These bio-nanomaterials combine the advantageous properties

39

of cellulose, namely the broad range of chemical modifications, renewability, and

40

sustainability with the specific attribute of nanosized materials. Among these nanosized

41

fibrils, nanofibrillar cellulose arouses much interest and holds promising applications in a

42

number of technical applications including nanofiller with outstanding reinforcing potential in

43

polymers’ matrixes (Boufi, Kaddami & Dufresne 2014, Chaker et al 2014, Besbes, Rei Vilar

44

& Boufi 2011) strength additive of papers (González et al 2011, 2012), in biomedical

45

applications such as drug delivery (Valo et al 2011), implantable scaffolds (Mathew et al

46

2011), highly porous cellulose adsorbents (Maatar, Alila & Boufi 2013)

47

NFC consists of long, flexible and entangled cellulose nanofibrils formed by alternating

48

crystalline and amorphous domains. It is composed of more or less individualized cellulose

49

microfibrils with lateral dimension in the order of 5 to 50 nm, depending on the production

50

and pretreatment process and length in the micrometer scale.

51

Currently, NFC is produced by disintegration of the cellulosic fibers under a high mechanical

52

shearing action using a high pressure homogenization (HPH), a micro-fluidization and even a

53

conventional high speed blender (Boufi & Gandini 2015, Chaker et al 2014) This intense

54

mechanical shear is necessary to overcome the strong interfibrillar hydrogen bonds and

55

break-up the cell walls to progressively release the nanosized microfibrils. In order to

56

facilitate the fibrillation process and reduce the energy demand during the disintegration

57

process, chemical fibers pretreatment is necessary, the most common of which has been

58

recourse to the introduction of charged groups within the fibers. This can be done via a

59

TEMPO-mediated oxidation (Saito et al 2006), periodate oxidation (Liimatainen et al 2012),

Ac ce p

te

d

M

an

us

cr

ip t

35

3 Page 3 of 29

carboxymetylation (Wågberg et 2008), sulfatation (Liimatainen et al 2013) and quaternization

61

to produce cationized NFC (Liimatainen et al 2014, Olszewska et al 2011).

62

In all of these chemical pretreatments, charged groups were introduced to generate repulsive

63

forces between microfibrils that contributed to loosen the microfibrils cohesion held by

64

hydrogen bonding. In fact, the origin of the repulsive forces is the osmotic pressure driven by

65

the difference in the ionic concentration between the interior and the exterior of the fibers.

66

Indeed, as the ionic groups are bounded to the fibrils surface and cannot move out, the

67

solution within the fibers’ pores might be regarded as separated from the external solution by

68

a semipermeable membrane that confines the ionic groups but gives free entry of water into

69

the fibres to reduce the osmotic differential pressure. The resulting fiber swelling contributed

70

to reduce the interfibrillar cohesion and facilitated the cell walls breakdown.

71

Only a limited number of publications have been concerned with cationic nanofibrillar

72

cellulose. The first work has been reported by Olszewska et al (Olszewska et al 2011 ) by

73

disintegration

74

trimethylammonium chloride in isopropanol. Ho et al prepared C-NFC by the conversion of

75

cellulose pulp with chlorocholine chloride in dimethylsulfoxyde (DMSO) followed by a

76

mechanical disintegration process (Ho et al 2011) . The C-NFC was used as a matrix to

77

embed layered silicates and prepare nanocomposite with improved water vapor barrier

78

properties. It was shown that C-NFC exhibited good interactions with negatively charged

79

layered silicaminerals in composite structures. Pei et al have used C-NFC to prepare

80

nanopapers with excellent mechanical properties which was used as an adsorbent for dye

81

removal from aqueous waste streams (Pei et al 2013).

82

In the present work, cationization was used as a pretreatment to facilitate the disintegration of

83

commercial softwood pulp into nanofibrillar cellulose via a high pressure homogenization.

84

Cationic nanofibrillar cellulose (C-NFC) with different contents of trimethylammonium

bleached

sulphite

pulp

cationized

with

N-(2-3-epoxypropyl)

Ac ce p

te

of

d

M

an

us

cr

ip t

60

4 Page 4 of 29

groups were produced and characterized using different techniques. A correlation between the

86

cationic content of the fibers and the efficiency of the nanofibrillation process was

87

demonstrated. The effect of the inclusion of C-NFC into a polymer matrix on the antibacterial

88

properties was also studied and compared to negatively charged NFC.

89 90

2. Experimental section

91

2.1. Materials

92

Commercial bleached Eucalyptus pulp (eucalyptus globulus) was used as starting material for

93

the preparation of NFC. The other chemical products were purchased from Sigma Aldrich and

94

used without purification.

95

2.2 Quaternization of cellulose fibers

96 97

The eucalyptus pulp was first soaked in distilled water for 12 h and subjected to a mechanical

98

disintegration using domestic warring blender to obtain a fiber suspension with 5wt %

99

concentration.

d

M

an

us

cr

ip t

85

te

The fiber suspension was then centrifuged to remove water and solvent

exchanged (twice) with dimethylacetamide (DMAC). Then a known amount of

101

glycidyltrimethylammonium chloride (GMA) was added to the fibers suspension in DMAC

102

(10wt %) and 0.2% of KOH was added as a catalyst. the reaction was carried out at 65°C for

103

8 hours. The reaction mixture was then centrifuged and washed five times with water to

104

completely remove any residual DMAC.

105

2.3 Trimethylammonium chloride content

106

The trimethylammonium chloride content (TMAC) was determined using conductometric

107

titration of chloride ions with AgNO3. The chloride counter ion is precipitated by reaction with

108

Ag+ until Briefly, about (100 mg) of Q-fibers were dispersed in 50 mL of water and titrated

109

with 5 mM silver nitrate (AgNO3). During the titration reaction, the conductivity remained

Ac ce p

100

5 Page 5 of 29

110

roughly constant until all the chloride is precipitated with AgNO3, then the conductivity grow

111

linearly.

112

2.4. Solid state Nuclear magnetic resonance (NMR)

113

Solid-state

114

All spectra were recorded by using a combination of cross-polarization, high power proton

115

decoupling and magic angle spinning (CP/MAS) methods.

116 117

2.5 Fibrillation process

118

NFC was prepared from the delignified pulp by pumping through a GEA Homogenizer

119

processor (NS1001L PANDA 2 K-GEA, Italy). The homogenization was conducted in two

120

steps. Firstly, the fibre suspension at a concentration of 1.5 wt% was passed three times at a

121

pressure of 300 bar (4350 Psi) until the suspension turned into a gel. Then, the fibrillation was

122

pursued by further five passes at a pressure of 600 bar (8700 Psi).

123

2.6. Yield in NFC:

124

A diluted gel (0.2% solid content) was centrifuged at 4500 rpm for 20 min to separate the

125

nanofibrillated material (in supernatant fraction) from non-fibrillated and partially fibrillated

126

fibres which sediment down. The sediment fraction was dried to constant weight at 90°C in a

127

halogen desiccator and the yield of the nanofibrillated fraction was calculated as such:

128

  weight of dried sediment   100 Yield %  1   (weight of diluted sample  %Sc) 

cr

ip t

C NMR experiments were performed on a Bruker AVANCE 400 spectrometer.

Ac ce p

te

d

M

an

us

13

129 130

%Sc= solid content of the diluted gel sample

131

The results represent the averages of three replications.

132

2.7. Nanocomposites processing:

133

Poly(vinyl alcohol) with 98–99 % hydrolysis degree was used as a matrix to prepare NFC

134

based nanocomposite films. PVA was first dissolved in water under stirring at 90 ° C to

135

obtain aqueous solutions of 10 wt% PVA concentrations. 6 Page 6 of 29

Nanocomposites were prepared via a solution casting method. The desired amount of NFC

137

and additional distilled water (to adjust the total solid content to 5% wt) were added to the

138

PVA solution and the mixture was mechanically stirred at room temperature for 1 h to ensure

139

an effective NFC dispersion. The nanocomposite solutions were degassed, then cast into

140

polystyrene petri dishes. The films were allowed to dry in the air at 23°C and 50% relative

141

humidity before being dried in an oven at 50°C for 4 h. before any mechanical testing, films

142

were stored in standard atmosphere (23 C, 50% relative humidity) for at least 7 days. The

143

nanocomposite films obtained were Translucent to transparent, according to the NFC content

144

with a thickness in the range of 200 to 400 µm.

145 146

2.8. X-ray diffraction analysis:

147

The crystalline degree of cellulose was calculated from an X-ray diffraction profile. X-ray

148

measurements were made on sample sheets cut in small pieces after the air drying of the

149

nanofibres suspension. Cu Kα radiation, generated with a Bruker AXS diffractometer (Bruker

150

AXS, Madison, WI at 30 kV, 100 mA. Sweeps of 5–50°2θ were made with a step size of 0.05

151

s and step time measurement of 10 s. The crystalline index (CrI) was calculated using the

152

diffraction intensities of the crystalline structure and that of the amorphous fraction according

153

to the method of Segal et al. (Segal, Creely, Martin & Conrad, 1959):

Ac ce p

te

d

M

an

us

cr

ip t

136

I  I  CrI %   002 am  100  I 002 

154 155

where I002 is the intensity of the crystalline peak at the maximum at 2θ between 22° and 23°

156

for cellulose I, and Iam is the intensity at the minimum at 2θ between 18° and 19° for cellulose

157

I.

158

2.9. Rheological measurements:

159

Rheological measurements were carried out using a controlled speed rotating rheometer

160

(Rheotec RS-30) operated with cone-and-plate geometries (cone angle: 4°, diameter: 3 cm) 7 Page 7 of 29

and a Peltier temperature controller. All the measurements were conducted at a constant

162

temperature of 25 ± 0.2 ◦C.

163

2.10. Antibacterial Activities

164

The antibacterial activity of the films was tested against Gram+ bacteria;Staphylococcus

165

aureus (ATCC 25923), Bacillus cereus (ATCC 14579) and Gram- bacteria (Enterococcus

166

faecalis (ATCC 29212). and (Escherichia coli (ATCC 25922), Salmonella enterica (ATCC

167

13883) and Salmonella typhi (ATCC 14028) using the disc diffusion method on an agar plate

168

with slight modification.

169

The diffusion method was performed using Luria−Bertani (LB) medium solid agar in Petri

170

dishes. The agar slurry on the test specimens was allowed to the gel at room temperature

171

followed by incubation at 37 °C for 18 h. After incubation, films were cut into 15× 15 mm

172

pieces, sterilized in autoclave at 121°C for 20 minutes, and kept in intimate contact with the

173

bacterial agar gel. After incubation during 24h at 37°C, samples were examined visually for

174

growth of bacteria in the area surrounding the film.

175 176

3. Results and discussion

177 178

3.1 Cationization of cellulose fibers

179 180

The cationization of cellulose with glycedyltrimethylammonium chloride is a well known

181

modification approach of cellulose fibers to generate a cationic ammonium site within the

182

cellulose fibers. These latter were incorporated onto cellulose through nucleophilic addition

183

of the hydroxyl groups of cellulose to the epoxy moiety of glycidyltrimethylammonium

184

chloride, as shown in scheme 1.

Ac ce p

te

d

M

an

us

cr

ip t

161

185 186 187 8 Page 8 of 29

GTA/fibers molGTA.mol- TMAC content CrI 1

Eucal-P*

-

AGU

(µmol.g-1)

Fibrillation yiel (%)

-

-

C-Eucal-210** 1

1.06

210

C-Eucal-350

0.3

0.21

300

75

C-Eucal-760

0.6

0.64

760

73

C-Eucal-1150

1

1.06

1140

76

75

8 35

ip t

samples

74 86

Table 1: TAMC content in the quaternized softwood pulp according to the amount of used GTA

192

and the fibrillation yield after homogenisation.

an

us

*pristine eucalyptus pulp. ** cationization performed without swelling in water

cr

188 189 190 191

In order to limit the hydrolysis of the epoxy ring during the addition reaction, DMAC was

195

used as a solvent instead of water, which is commonly used for this reaction. Moreover, to

196

ensure efficient swelling of the fibers and good accessibility of the reagent to the inner pore,

197

the fibers were first soaked in water for several hours and solvent exchanged with DMAC.

198

Catalytic amount of NaOH was also added to convert alcohol groups of cellulose into more

199

reactive alcoholate groups.

200

Evidence of the occurrence of the addition reaction was confirmed by 13C CP/MAS (Figure 1)

201

showing in addition to the cellulose backbone between 60 and 110 ppm, the typical signal of

202

the (CH3)3N+ moiety appearing around 55 ppm. The resonances of the cellulose skeleton

203

appeared at 105 (C-1), 88 (crystalline C-4), 84 (amorphous C-4), 75, 72 (C-2, C-3, C-5), and

204

64 (C-6) ppm. Based on the peaks integration, the content of the trimethylammonium chloride

205

is determined with accuracy and found to be in good agreement with conductometric titration

206

of chloride ions using AgNO3. By changing the ratio of glycedyltrimethylammonium

207

chloride/cellulose fibers, three samples of cationized fibers with various trimethylammonium

208

contents

209

given in Table 1. It is worth noting that, when the reaction was performed in DMAC without

Ac ce p

te

d

M

193 194

were prepared. The content of TMAC according to the GMC/Cellulose ratio is

9 Page 9 of 29

swelling the fibers in water, the cationic content of the fibers was about five fold lower than

211

that obtained with the swelling treatment (C-NFC-210) (see Table 1). This difference might

212

be explained by the lower accessibility of the hydroxyl groups within the cellulose fibers.

213

Indeed, delignified cellulose fibers exhibit a multiscale hierarchical structure and are a porous

214

material which can absorb water and other molecular components. This internal porosity,

215

created during the lignin removal, opens the cell wall and allows the fibers to expanse in

216

water, making them accessible to a wide range of molecules with different size (Thompson,

217

Chen & Grethlein 1992). Since cationization is presently adopted to facilitate the fibrillation

218

process, it is crucial that the reaction occurred inside the fibers on the surface of the accessible

219

microfibrils.

M

an

us

cr

ip t

210

C ellu lose M icrof ib ril

Ac ce p

te

d

C2,3,5

O

CH2

CH

CH2

N(CH3)3+

OH

C1

C6

C4 N(CH3)3+

220 221

Figure 1 : CP-MAS 13C NMR spectra of quaternized fibers with TMAC content of 1140

222

µmol.g-1. (C1 to C6 stands for carbones of the glucosic unit of cellulose) 10 Page 10 of 29

223 Analysis of the crystalline structure of the fibers by X-ray diffraction, revealed no evolution

225

of the cellulose I polymorph, nor in the crystallinity degree after the cationization treatment

226

(see supplementary material), which indicates that the cationization reaction occurred mainly

227

on the surfaces of the cellulose microbrils. All the samples showed a typical diffraction

228

pattern of cellulose I, with the typical diffraction peaks at 2 16, 22.6°. The CrI remained

229

roughly constant around 75% irrespective of the TMAC content.

230 231

3.2 Nanofibrillation behaviour of the cationized fibers

232

To investigate how the cationic content affected the fibrillation behavior of the fibers, a high

233

pressure homogenization was progressively carried out, starting with several passes at 300 bar

234

until the suspension turned into a gel. Then, it was pursued at 600 bar to further promote the

235

breakdown of the cell walls. During the disintegration process, samples were withdrawn at

236

regular passes and the NFC suspension was analyzed in terms of its nanofibrillation yield,

237

transparency and rheological behavior.

238

A strong dependence of the fibrillation yield with the TMAC content of the fibers was noted

239

(Table 1, Figure 2 supplementary material), with a huge enhancement in the fibrillation

240

efficiency as the TMAC content increased. The fibrillation yield reached about 35, 70 and

241

85% as the cationicity grew from 350, to 750 and 1160 µmol.g-1, respectively. In the absence

242

of any pretreatment and adopting the same disintegration mode, the fibers failed to be

243

converted into NFC and multiple clogging occurred during high pressure homogenization.

244

This phenomenon was not observed in the presence of cationized treated fibers.

245

How to explain the strong enhancement of the nanofibrillation efficiency with the TMAC

246

content? In fact, cellulose is a hierarchical material formed of elementary fibrils aggregated

247

into larger microfibrils 5–50 nm in diameter and several microns in length, sticking together

248

in a spiral manner within a multiple-layer-composite to form the cellulose fibers. Upon

Ac ce p

te

d

M

an

us

cr

ip t

224

11 Page 11 of 29

immersion in water, the fibers underwent expansion as a result of water diffusion inside the

250

fibers microspores. Then, after solvent exchange and addition of GMA, the accessible

251

microfibrils, namely those in the amorphous domain, were cationized with GMA, resulting in

252

the generation of quaternary trimethylammonium groups bounded on the surface of

253

microfibrils. The presence of these ionic groups within the cell wall resulted in an osmotic

254

swelling of the fibers that reduced the hydrogen bonding among the microfibrils by increasing

255

the gap between neighbouring fibrils as schematically illustrated in scheme 1. The

256

proportionality between the osmotic pressure and the content of ionic groups

257

Posm  RT (

258

and V is the liquid volume of the fibre wall (m3/g) accounted for the enhancement in the

259

fibrillation yield as the content in TMC is going up.

us

cr

ip t

249

260

Ac ce p

OH OH

te

OH OH

d

M

an

N ion ) ( Posm = where Posm (N/m2), Nion is the moles of charged groups (moles/g), V

OH O CH2 CH CH2 N(CH3)3+ OH

Cl-

OH

Microfibril

Elementary fibril

261 262

P >P + osm+ ext +

+ +

+

+ +

Electrostatic repulsion

+

+

+

Osmotic pressure n Posm  c RT V

263 264

Scheme 1: schematic illustration of how the fibers quaternization enhanced the

265

nanofibrillation process.

266 12 Page 12 of 29

The transparency degree of the C-NFC after the homogenization process was shown to be

268

greatly affected by the TMAC content of the fibers. As shown in Figure 2, at a TMAC content

269

of 350 µmol.g-1, the C-NFC gel is quite opaque and turned to a translucid gel as the TMAC

270

increased over 750 µmol.g-1. This change in the optical transparency is further highlighted by

271

the transmittance measurement of a diluted suspension of C-NFC. The transmittance level for

272

C-NFC-350 did not exceed 30% and grow to 65-70% and 72-80% for C-NFC-750 and C-

273

NFC-1160 respectively, which is indicative of the low fraction of micro-sized particles

274

responsible for light scattering. The evolution in the optical transparency is explained by the

275

change in the fibrillation yield and the dispersion of the cellulose nanofibrils in water. The

276

enhancement in the transmittance with increasing content of ionic groups is indicative of a

277

decrease in the amount of non-fibrillated and partially fibrillated fractions responsible for a

278

light-scattering phenomenon. In addition, the increase in the TMAC groups is expected to

279

enhance the colloidal stability of the cellulose nanofibrils through electrostatic stabilization

280

brought by the presence of cationic groups on the surface of the nanofibrils. This stabilization

281

prevents the likelihood tendency of nanosized fibrils to self-aggregate through hydrogen

282

interaction. The aggregation risk of nanosized particles is another phenomenon likely to affect

283

the transparency degree of the suspension.

284

The transmittance spectra of C-NFC also showed also a strong wavelength-dependence with

285

lower transmittance at shorter wavelength, namely with TMA content of 750 and1160

286

µmol.g-1. This phenomenon is explained by the dependence of scattering intensity to the

287

fourth power of wavelength ( I  1 / 4 ) for nanosized particles with square section lower than

288

50 nm.

Ac ce p

te

d

M

an

us

cr

ip t

267

289

13 Page 13 of 29

A 1140

300

760

ip t

760

300

1140

cr

290 90 80

1140 µmol.g-1

us

B

760 µmol.g-1

an

60 50

300 µmol.g-1

40

M

Transmittance (%)

70

30

291

500

600

700

800

Wavelength (nm)

Ac ce p

0 400

te

10

d

20

292

Figure 2. (A) Visual aspect of Q-NFC gel at a solid content of 1 wt% with different TMAC

293

amount, and (B) Visible transmittance spectra of the Q-NFC suspension for different TMAC

294

content after 5 passes at 300 bar and 5 additional passes at 600 bar

295 296

The AFM images of C-NFC samples deposited on wafer substrate and dried at room

297

temperature are shown in Figure 3. these observations were relative to C-NFC samples with a

298

TMAC content of 750 and 1160 %mol.g-1 and produced via high pressure homogenization

299

following the same pressure sequence (5 passes at 300 bar followed by 5 additional passes at

300

600 Bar). Both samples showed a network of entangled individual nanofibrils having typical

301

width of 20–40 nm and lengths in the micrometer range. The length evaluation of the fibril is 14 Page 14 of 29

qualitative as the long fibrils formed an entangled network. Besides, it is hard to evaluate the

303

entire length of the fibrils accurately.

304

The fibril size, estimated from the height of the fibrils directly attached to the wafer, was

305

found to be between 3 and 6 nm for the individual non-aggregated fibrils. The NFC produced

306

from fibers with a TMAC content of 1160 µmol.g-1 seems to exhibit lower square section than

307

that with lower TMAC content. The difference between the thickness of the fibrils and their

308

width is likely to be due to the dimensions and geometry effect of the AFM tip that led to

309

surface feature broadening. This latter effect became obvious when the surface features have

310

the same or smaller size than the AFM tip. Under this condition, the tip will not be able to

311

correctly draw the profile contours giving rise to the so called convolution effects with artifact

312

appearance results, and the height should give more reliable measurement of NFC width.

313

From the AFM analysis, it can be confirmed that cationized celluloses were efficiently

314

homogenized to individual nanofibrils having lateral dimensions lower than 8 nm. Further

315

confirmation of the nanoscale dimension of C-NFC was given by FE-SEM observation

316

(Figure 4) with a larger area view of C-NFC, and where the partially fibrillated material with

317

width within 100 nm was clearly shown.

319

cr

us

an

M

d

te

Ac ce p

318

ip t

302

15 Page 15 of 29

A

4.2 nm

3.5 nm

us

cr

ip t

2.5 nm

320

an

321

6.4 nm 4 nm

322 323

Ac ce p

te

d

M

B

324

Figure 3: AFM height images of Q-NFC with TMAC content of (A) 1140 and (B) 760 µmol.g-

325

1, and the corresponding height profile analysis (the arrow marks the points the points used

326

for the measurements of the height profile

327

16 Page 16 of 29

90 nm 10 nm

us

cr

ip t

Partially fibrillated material

Figure 4 : FE-SEM images of C-NFC produced from quaternized fibers with TMAC content of 350 µmol.g-1. (C-NFC-350)

333

3.3 Colloidal properties of C-NFC

334 335

The colloidal stability and the rheological behavior of NFC in water are highly dependant on

336

the zeta-potential of the cellulose nanofibrils. The electrokinetic properties and the rheological

337

behavior of C-NFC were studied according to the TMAC content. The evolution of the zeta-

338

potential versus pH for the C-NFC suspension is shown in Figure 5. For the different contents

339

in TMAC, the nanofibrils exhibited a positive ζ-potential over the whole range of pH.

340

However, the magnitude of the ζ-potential is shifted to higher value as the TMAC content

341

increased and showed a pH dependence between pH 5 and 11. The former effect is expected

342

since a higher content in TAMC led to an increase in the surface charge density of the fibrils.

343

The value of ζ-potential exceeding +30 mV for C-NFC with a TMA content of 760 and 1160

344

µmol.g-1 accounts for the good dispersion stability of the NFC suspension in water as attested

345

by the individualization of the nanofibrils in AFM observations and good transparency of the

346

C-NFC suspension, even at a solid content over 1wt.% .

Ac ce p

te

d

M

an

328 329 330 331 332

17 Page 17 of 29

40 1140 µmol.g

-1

30

760 µmol.g

-1

ip t

25 20 300 µmol.g

-1

cr

Zeta-potential (mV)

35

10 2

3

4

5

6

7 pH

348

8

9

10

11

an

347

us

15

Figure 5: Change in the zeta-potential versus pH of Q-NFC with different TMAC content.

M

349

d

1000

te

Ac ce p

Viscosity (Pa.s)

100

1140 µmol.g-1

10

760 µmol.g-1

1

300 µmol.g-1

0.1

0

2

4

6

8

10

12

Shear rate (s-1)

350 351

Figure 6. Viscosity as a function of shear rate of the Q-NFC gel at 1% solid content with

352

different TMAC content.

353

NFC suspension is known to exhibit a gel-like aspect as the solid content exceeds a critical

354

threshold in the range of 0.1-5 wt%, depending on the origin and the morphological features 18 Page 18 of 29

of the cellulose nanofibrils. The evolution of the viscosity vs. shear rate of C-NFC at different

356

TMAC contents is shown in Figure 6. All of the C-NFC suspensions exhibited a strong shear

357

thinning behavior, namely when the TMAC content exceeds 750 µmol.g-1, as attested by the

358

intense decrease in the viscosity as the shear rate increases. For instance, the viscosity of the

359

C-NFC gel with a cationicity of 1160 µmol.g-1, falls down from 725 to about 35 Pa.s, as the

360

shear rate goes from 0.4 to 5 s-1. This behaviour, typical of most NFC suspensions in water, is

361

explained by the hydrodynamic properties of charged cellulose nanofibrils with a high aspect

362

ratio. Indeed, , the NFCs flexibility along with their strong tendency to self interact through

363

hydrogen bonding led to the formation of coiled network formed by entangled cellulose

364

fibrils. Upon the application of a shear rate the network is progressively broken down, the

365

elementary nanofibrils start to be oriented in the shear direction leading to their

366

disentanglement. Consequently, the flow resistance is lowered and the viscosity is decreased.

367

Once in rest, the nanofibrils network is reformed and the viscosity rebuilt again. The shift in

368

the viscosity toward higher value as the cationic content increases is mainly due to the

369

enhancement in the fibrillation extent- as the TMAC content exceeded 700 µmol.g-1.

370 371

3.4. Antibacterial properties.

372 373

To evaluate the antibacterial activity of the C-NFC when incorporated in a polymer matrix,

374

the diffusion method using Gram+ and Gram- bacteria was adopted. Pure PVA matrix and

375

nanocomposite films prepared from negatively charged A-NFC and PVA matrix was used as

376

a reference for the sake of comparison.

377

Figure 7 presents the picture of the diffusion assay resulting from pure PVA films, C-NFC-

378

PVA and A-NFC-PVA nancomposite films with 3 and 7% nanofiller loading. Pure PVA film

379

did not show any inhibitory effects on .all the bacterial tested. The same tendency was

380

observed for nanocomposite films prepared with A-NFC, with colonies growing of BC and E-

Ac ce p

te

d

M

an

us

cr

ip t

355

19 Page 19 of 29

colis growing on the agar plates as well as on the surface of the A-NFC-PVA films. On the

382

other hand, for all nanocomposite films prepared with C-NFC, bacteria increased on the agar

383

plates around the films, but not on top or under the films, and no trace of colonisation of the

384

film was observed. In the presence of Staphylococcus aureus and Salmonella enterica, better

385

antibacterial activity was observed as attested by the appearance of inhibition zone around the

386

film. The absence of a clear and large inhibition zone is probably due to the immobilization

387

of antibacterial agent (C-NFC) on the polymer matrix restricting the diffusion of the

388

antibacterial agent to areas close to the specimen. The antibacterial activity of C-NFC-PVA

389

seems to be independent on the NFC loading over the range between 3 to 10wt. %. These

390

results clearly demonstrate the trivial role of C-NFC in preventing bacteria proliferation when

391

incorporation into a polymer matrix at a content as low as 4wt%. The efficient antibacterial

392

activity of C-NFC against a board range of bacteria is likely due to the presence of quaternary

393

ammonium groups on the surface of the C-NFC. These groups are known to be one of the

394

most useful antiseptics and disinfectants used in wide range of applications in cosmetic

395

products, clinical purposes, textiles and active packaging. The mechanism proposed to

396

account for these strong antibacterial properties is the interaction with the lipids and proteins

397

that compose the cytoplasmic membrane damaging the outer layer of bacterial disturbing their

398

metabolism (Denyer 1995).

cr

us

an

M

d

te

Ac ce p

399

ip t

381

20 Page 20 of 29

399 a

C-NFC

b

c

1 401 4 3

4

403

405 406

Bacillus cereus

C-NFC Staphylococcus aureus

C-NFC

A-NFC

d

an

408 409 410 Escherichia coli

M

Bacillus cereus

412 413

418

A-NFC

te

d

Figure 7. The antibacterial activity of control NFC-PVA nanocomposite film prepared from quationized NFC C-NFC (a, b and c) and from anionic A-NFC (d, e) tested against Gram- and Gram+ bacteria: (1) neat PVA matrix, (2) 3% NFC, (3) 5% NFC and (4) 7% NFC.

Ac ce p

414 415 416 417

C-NFC Salmonella enterica

e

407

411

3

3

us

404

4

ip t

402

1

1

cr

400

C-NFC

419

3.5 Reinforcing potential

420

In addition to the antibacterial effect imparted by the inclusion of C-NFC into a polymer

421

matrix, a reinforcing effect is also expected by the addition of NFC. To investigate the

422

reinforcing potential of the C-NFC, nanocomposite films based on PVA matrix and different

423

nanofiller loadings were prepared by solvent-casting and were analysed by DMA. For the

424

sake of comparison, similar films were also prepared using nanofibrillar cellulose bearing

425

carboxylic groups (A-NFC). The temperature dependence of the storage tensile modulus, E΄ ,

426

as a function of temperature at 1 Hz for PVA-NFC (C-NFC and A-NFC) nanocomposite with

427

different NFC are reported in the supplementary material. 21 Page 21 of 29

30

(A)

At 15% A-NFC 27-fold matrix

25

E'c/E'm.100

20

ip t

15 10

cr

At 15% Q-NFC 11 fold matrix

5

0

1

2

3

4

5

6

7

us

0 8

9 10 11 12 13 14 15 16

429 (B)

an

NFC content (wt%)

428

Cationized Cellulose nanofibril

OH

OH

OH CH2

CH2 CH O

OH

N+

H3C OH CH3

OH

OH

CH3 N+ H3C OH CH3

OH

d

OH

OH OH CH CH3 CH2 CH2 O

M

OH

OH

te

Cationized Cellulose nanofibril

430 431

OH

Ac ce p

OH

OH

OH

OH

OH

OH

OH

OH

PVA matrix

(C)

Carboxylated Cellulose nanofibril

OH

OH

O

OH

HO

C

O

OH

OH

OH

C

OH

OH O

OH

OH

HO

C

O

OH

OH C

OH

OH

Carboxylated Cellulose nanofibril

OH

OH

COOH

OH

OH

OH OH

432 433

OH OH

PVA matrix

22 Page 22 of 29

Figure 8. (A) The relative storage modulus versus NFC content at 70 °C for nanocomposite films prepared from C-NFC and A-NFC and schematic illustration of the hydrogen potential interaction among (B) C-NFC, and(C) A-NFC, when embedded into a polymer matrix.

438

At a temperature lower than the glass transition, the storage modulus undergoes only a slight

439

enhancement with respect to the unfilled matrix. For instance, at 15% NFC E΄ is about two

440

fold higher. Besides, at 20°C the nanocomposites containing 15 wt% C-NFC and A-NFC

441

exhibit a storage modulus around 1.2 and 1.4 GPa, which represents a 2.8 and 3.3-fold

442

enhancement, respectively, over that of the neat matrix (0.42 GPa). This behavior, common to

443

nanocomposite materials based on nanosized cellulose particles, is explained by the low

444

difference between the modulus of the glassy matrix (close to 1 GPa) and that of the

445

nanofiller network (being in the range of 10–15 GPa).

446

Above the glass transition, however, the storage modulus is more sensitive to the presence of

447

the nanofiller and increased significantly with NFC addition, which is in line with the well-

448

known reinforcing effect of nanocellulose based nanofiller.(Boufi 2014)

449

The stiffening effect imparted by the addition of NFC can be expressed by the evolution of

450

the relative modulus Er 

451

and unfilled matrix respectively measured in the rubbery region taken here at 70°C), versus

452

the NFC content (Figure 8A). The continuous rise in the relative modulus clearly confirms the

453

reinforcing potential of the NFC. However, the extent of the reinforcing potential is different

454

from cationic to anionic NFC. At the same nanofiller loading, the modulus enhancement is

455

lower for C-NFC than that observed for A-NFC. This is namely observed when the NFC

456

content exceeded 2%. For instance, at 15 wt %, the modulus is about 27 and 11 times higher

457

for samples prepared with A-NFC and C-NFC respectively.

458

To explain the lower reinforcing potential of C-NFC compared to that of A-NFC, it is worth

459

noting that the unusual enhancement of the stiffness and strength in the nanocellulose

te

d

M

an

us

cr

ip t

434 435 436 437

Ac ce p

E 'c (with Ec and Em are the storage modulus of the nanocomposite E 'm

23 Page 23 of 29

reinforced nanocomposites arouse mainly from the setup of an entangled interconnected

461

network held-up by strong hydrogen bonding. The strength of the NFC network is dependant

462

on the aspect ratio of the cellulose nanofibrils and their capacity to interact through hydrogen

463

bonding. Presently, we infer that the presence of a relatively high surface density of appended

464

2-hydroxypropyltrimethylammonium chloride moiety reduced the potential of hydrogen

465

bonding among neighbouring nanofibrils and weakened the network cohesion. This can be

466

deduced from the following analysis. Indeed, assuming that the square section of the

467

nanofibrils is 5 nm and that the average values between cellulose chains within a transversal

468

cross section ranges from 0.54 to 0.61 nm, then for a content of 1100 µmol.g-1, the surface

469

substitution degree should be DSS=DS/ratio of surface chains=0.19/0.4=0.47 . This means

470

that

471

hydroxypropyltrimethylammonium

472

trimethylammonium group to share hydrogen bonding with hydroxyl groups, the interaction

473

among entangled cellulose nanofibrils network is expected to weaken due to the shielding

474

effect imparted by the trimethylammonium group. For A-NFC, the presence of carboxylic

475

groups on the surface of the nanofibrils will, on the contrary, favour hydrogen interaction

476

among NFC. A schematic illustration of this phenomenon is given in Figure 8Aand B.

477

Though, the production of C-NFC via the quaternization of cellulose fibers as a pretreatment

478

was already reported in several paper as noted in the introduction (Olszewska et al 2011, Ho

479

et al 2011 and Pei et al 2013), the main finding of the present work is twofold; correlate the

480

fibrillation efficiency with the content of cationic groups, and highlight the great potential of

481

C-NFC to enhance the antibacterial resistance of the material when included as additive into a

482

polymer matrix. The potential usefulness of C-NFC gel as wound healing might be also

483

considered. It follows that the quaterization of NFC as both a pretreatment and method for

each

two

anhydroglucosic

M

of

chloride

units

moiety.

is

Given

linked the

to lack

the

2-

of

the

Ac ce p

te

d

one

an

us

cr

ip t

460

24 Page 24 of 29

surface modification open the way for usefulness of C-NFC in biomedical application. This

485

work is under investigation.

486

4. Conclusion

487

Cationized NFC from eucalyptus pulps was prepared by cationization of the cellulose fibers

488

with GTMAC in DMF and mechanical disintegration using a high pressure homogenization.

489

What can be concluded from the present study is that:

cr

490

ip t

484

1- The cationization of the bleached cellulose fibers is an efficient approach to facilitate the breakup of the cell wall and the release of cellulose nanofibrils.

us

491

2- The fibrillation efficiency is meaningfully enhanced as the TMAC content exceeded

493

700 µmol.g-1. Over this content, thick transparent gel was obtained with good

494

colloidal stability for long-term storage.

M

495

an

492

3- AFM observation confirmed the nanosized scale of C-NFC with a square section in the range of 5 nm.

d

496

4- The inclusion of C-NFC within a polymer matrix strongly enhanced the antibacterial

498

activity of the nanocomposite films by imparting a reinforcing effect. These

499

nanocomposites films could be useful in active packaging material for food

501 502

Ac ce p

500

te

497

preservation. These results clearly suggested the great potential of cationic NFC as reinforcement additive for polymer with effective antibacterial activity.

503

References

504

Besbes I., Rei Vilar M., & Boufi S.(2011) Nanofibrillated cellulose from alfa, eucalyptus and

505

pine fibres: preparation, characteristics and reinforcing potential. Carbohyd. Polym. 86, 1198-

506

1206.

507 508

Boufi S. Chapter 14 in . Biomass and Bioenergy. Khalid Rehman Hakeem,Mohammad

509

Jawaid,Umer Rashid. Springer 2014, pp 267-305.

510 25 Page 25 of 29

Boufi S., Kaddami H.,& Dufresne A. (2014) Mechanical performance and transparency of

512

nanocellulose reinforced polymer nanocomposites. Macromol. Mater. Eng. 299, 560–568

513 514

Boufi S.,& Gandini A.. Triticale crop residue: a cheap material for high performance

515

nanofibrillated cellulose, RSC Adv., 2015, 5, 3141–3151

516 517

Chaker A., Mutjé P., Rei Vilar M.,& Boufi S. (2014) Agriculture crop residues as a source for

518

the production of nanofibrillated cellulose with low energy demand, Cellulose 21, 4247–4259

519

cr

ip t

511

Chaker A., Mutje P., Vilaseca F.,& Boufi S. (2014) Reinforcing potential of nanofibrillated

521

cellulose from nonwoody plants. Polym. Compos. 34, 1999-2007.

522

Denyer, S.P. (1995) Mechanisms of action of antibacterial biocides. Int. Biodeterior.

523

Biodegradation 36, 227–245

an

us

520

524

González I., Boufi S., Pèlach M. A., Alcalà M., Vilaseca F.,& Mutjé P. (2012)Nanofibrillated

526

cellulose as paper additive in bleached hardwood pulps. Bioresources 7, 5167-5180.

527 528

González I., Vilaseca F., Alcalá M., Pèlach M., Boufi S.,& Mutjé P. (2013) Effect of the

529

combination of biobeating and NFC on the physico-mechanical properties of paper. Cellulose,

530

20 (3), 1425-1435.

531 532

Ho T. T. T., Zimmermann T., Hauert R.,& Caseri W.,(2011) Preparation and characterization

533

of cationic nanofibrillated cellulose from etherification and high-shear disintegration

534

processes, Cellulose 18, 1391–1406

535 536

Kalia S., Boufi S., Celli A., & Kango S. (2014) Nanofibrillated cellulose: surface

537

modification and potential applications. Colloid. Polym. Sci. 292, 5–31

d

te

Ac ce p

538

M

525

539

Khalil H.P.S.A, Bhat A.H,& Yusra A.F.I,(2012) Green composites from sustainable cellulose

540

nanofibrils: A review, Carbohydrate Polymers 87 (2), 963-979

541 542

Liimatainen H., Suopajärvia T., Sirviöa J., Hormib O.,& Niinimäki J. (2014) Fabrication of

543

cationic cellulosic nanofibrils through aqueousquaternization pretreatment and their use in

544

colloid aggregation, Carbohydrate Polymers 103, 187– 192

545

26 Page 26 of 29

546

Liimatainen H., Visanko M., J Sirvio¨ J., Hormi O.,& Niinima¨ki J., (2013) Sulfonated

547

cellulose nanofibrils obtained from wood pulp through regioselective oxidative bisulfite pre-

548

treatment, Cellulose 20,741–749

549 Liimatainen H., Visanko M., Sirvio¨ J., Hormi O., Niinima¨ki J (2012) Enhancement of the

551

nanofibrillation of wood cellulose through sequential periodate–chlorite oxidation.

552

Biomacromolecules 13, 1592–1597

553 554

Maatar W., Alila S.,& Boufi S. (2013) Cellulose based organogel as an adsorbent for

555

dissolved organic compounds. Ind Crop Prod 49, 33–42

556 557

Mathew A.P, Oksman K., Pierron D.,& Harmand M.F. (2012) Fibrous cellulose

558

nanocomposite scaffolds prepared by partial dissolution for potential use as ligament or

559

tendon substitutes. Carbohydr Polym 87:2291–2298

560

Olszewska A., Eronen P., Johansson L.S., Malho J.M., Ankerfors M., Lindstrom T.,

561

Ruokolainen J., Laine J.,& Osterberg M. (2011) The behaviour of cationic NanoFibrillar

562

Cellulose in aqueous media, Cellulose 18, 1213–1226

563 564

Pei A., Butchosa N., Berglund L.A., Zhou Q.(2013) Surface quaternized cellulose nanofibrils

565

with high water absorbency and adsorption capacity for anionic dyes, Soft Matter, 9, 2047–

566

2055

567

Saito, T., Nishiyama, Y., Putaux, J. L., Vignon, M., & Isogai, A. (2006). Homogeneous

568

suspensions of individualized microfibrils from TEMPO-catalyzed oxidation of

569

native cellulose. Biomacromolecules, 7, 1687–1691.

cr

us

an

M

d

te

Ac ce p

570

ip t

550

571

Thompson D. N., Chen H-C,& Grethlein H. E., (1992) Comparison of Pretreatment Methods

572

on the Basis of Available Surface Area Bioresource Technology 39, 155-163

573 574

Valo H., Kovalainen M., Laaksonen P, Hakkinen M, Auriola S, Peltonen L, Linder M,

575

Jarvinen K, Hirvonen J,& Laaksonen T (2011) Immobilization of protein-coated drug

576

nanoparticles in nanofibrillar cellulose matrices-enhanced stability and release. J Control

577

Release 156:390–397

578

Wågberg L, Decher G, Norgren M, Lindstro¨m T, Ankerfors M,& Axna¨s K (2008) The

579

build-up of polyelectrolyte multilayers of microfibrillated cellulose and cationic

580

polyelectrolytes. 27 Page 27 of 29

581

Langmuir 24:784–795

Ac ce p

te

d

M

an

us

cr

ip t

582

28 Page 28 of 29

583

Figure captions

584 Figure 1 : CP-MAS 13C NMR spectra of quaternized fibers with TMAC content of 1140

586

µmol.g-1. (C1 to C6 stands for carbones of the glucosic unit of cellulose)

587

Scheme 1: schematic illustration of how the fibers quaternization enhanced the

588

nanofibrillation process.

589

Figure 2. (A) Visual aspect of Q-NFC gel at a solid content of 1 wt% with different TMAC

590

amount, and (B) Visible transmittance spectra of the Q-NFC suspension for different TMAC

591

content after 5 passes at 300 bar and 5 additional passes at 600 bar

592

Figure 3: AFM height images of Q-NFC with TMAC content of (A) 1140 and (B) 760

593

µmol.g-1, and the corresponding height profile analysis (the arrow marks the points the points

594

used for the measurements of the height profile

595

Figure 4 : Change in the zeta-potential versus pH of Q-NFC with different TMAC content.

596

Figure 5. Viscosity as a function of shear rate of the Q-NFC gel at 1% solid content with

597

different TMAC content.

598 599 600 601 602 603 604 605 606

Figure 6. The antibacterial activity of control NFC-PVA nanocomposite film prepared from C-NFC (a, b and c) and from A-NFCtested against Gram- and Gram+ bacteria: (1) neat PVA matrix, (2) 3% NFC, (3) 5% NFC and (4) 7% NFC.

te

d

M

an

us

cr

ip t

585

Ac ce p

Figure 7. the relative storage modulus versus NFC content at 70 °C for nanocomposite films prepared from Q-NFC and A-NFC, and schematic illustration of the hydrogen potential interaction among (A) C-NFC, and A-NFC, when embedded into a polymer matrix

29 Page 29 of 29