Analysis of the porous architecture and properties ... - ACS Publications

Subscriber access provided by UNIV OF DURHAM

Article

Analysis of the porous architecture and properties of anisotropic nanocellulose foams - A novel approach to assess the quality of cellulose nanofibrils (CNFs) Konstantin Kriechbaum, Pierre Munier, Varvara Apostolopoulou-Kalkavoura, and Nathalie Lavoine ACS Sustainable Chem. Eng., Just Accepted Manuscript • DOI: 10.1021/acssuschemeng.8b02278 • Publication Date (Web): 18 Jul 2018 Downloaded from http://pubs.acs.org on July 19, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sustainable Chemistry & Engineering

1

Analysis of the porous architecture and properties of

2

anisotropic nanocellulose foams - A novel approach

3

to assess the quality of cellulose nanofibrils (CNFs)

4

Konstantin Kriechbaum,† Pierre Munier,† Varvara Apostolopoulou-Kalkavoura,†

5

and Nathalie Lavoine,†,‡,*

6 7 8

†

Department of Materials and Environmental Chemistry, Stockholm University, Arrhenius Laboratory, Svante Arrhenius väg 16 C, 106 91 Stockholm, Sweden

*Corresponding author: Dr. Nathalie Lavoine; Tel: +1 919 513 3235; [email protected]

9

ACS Paragon Plus Environment

1

ACS Sustainable Chemistry & Engineering 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 37

10

ABSTRACT

11

Cellulose nanofibrils (CNFs) are a unique nanomaterial because of their abundant, renewable, and

12

biocompatible origin. Compared with synthetic nanoparticles, CNFs are commonly produced from

13

cellulose fibers (e.g. wood pulp) by repetitive high-shear mechanical disintegration. Yet, this

14

process is still highly demanding in energy and costly, slowing down the large-scale production

15

and commercialization of CNFs. Reducing the energy consumption during fibers fibrillation

16

without using any chemical or enzymatic pretreatments while sustaining the CNF quality, is

17

challenging. Here, we show that the anisotropic properties of the CNF foams are directly connected

18

to the degree of nanofibrillation of the cellulose fibers. CNFs were produced from wood pulps

19

using a grinder at increasing specific energy consumptions. The anisotropic CNF foams were made

20

by directional ice templating. The porous architecture, the compressive behavior of the foams and

21

the CNF alignment in the foam cell walls were correlated to the degree of fibrillation. A particular

22

value of specific energy consumption was identified in respect to the highest obtained foam

23

properties and CNF alignment. This value indicated that the optimal degree of fibrillation, and thus

24

CNF quality, was achieved for the studied cellulose pulp. Our approach is a straightforward tool

25

to evaluate the CNF quality, and a promising method for the benchmarking of different CNF

26

grades.

27

KEYWORDS: Nanocellulose; Cellulose nanofibril; Energy consumption; Degree of fibrillation;

28

Ice templating; Anisotropic foam;

29


2

Page 3 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

30


INTRODUCTION

31

As an alternative to fossil-based materials, nanocellulose is the focus of intense research

32

activities. Cellulose nanofibrils (CNFs) are one type of nanocellulose, with diameters ranging from

33

5 to 30 nm and aspect ratios (Length/Diameter) greater than 50.1 Besides being renewable and

34

biodegradable, CNFs combine low density with high mechanical strength, chemical inertness and

35

versatility.2 The use of CNFs has been considered for a wide range of applications including

36

composite reinforcement,3–6 food packaging,7,8 coatings,9 insulation,10–12 biomedical13–15 and

37

electronic16 applications. CNFs can be produced from any cellulose sources by successive

38

mechanical disintegration which exposes the cellulose material to high shearing forces, thus

39

promoting the separation of cellulose fibers into CNFs. Cost-efficient production is, however, still

40

challenging due to high-energy consumption and processing methods which are hard to scale up.17

41

To lower the energy consumption during disintegration, enzymatic18 or chemical19,20

42

pretreatments are commonly performed prior to the mechanical treatment. However, the life cycle

43

assessment of CNFs production shows that chemical pretreatments pose high environmental

44

impacts due to e.g. the use of solvents such as ethanol or isopropanol.21 Enzymatic pretreatment

45

and pure mechanical disintegration have similar low environmental impacts. Yet, the use of

46

phosphate buffer and high amount of washing water contribute strongly to the terrestrial

47

acidification and water depletion when using enzymes.22 Lowering the environmental impact of

48

the no-pretreatment route can be achieved by reducing the electricity usage in selecting the right

49

equipment and running it in a sustainable manner.21,22 Among the various mechanical treatments23–

50

25

51

of a grinder shows to this end various advantages: (i) the highest capacity for CNFs production,

52

(ii) the absence of clogging caused by large fibers promoting (iii) the production of CNFs without

reported to produce CNFs such as homogenization,26 microfluidization6 and grinding,27 the use


3


Page 4 of 37

53

performing any pretreatment, and (iv) the best energy efficiency.17 Comparing the production

54

efficiency of CNFs obtained with a grinder only is, however, more complex, as no standard

55

protocol has been set up for a sustainable and energy-efficient production. Many different

56

parameters influence the fibrillation outcome during grinding, such as the gap clearance between

57

the grinding discs, the rotational speed of the lower disc, the number of grinding cycles, and the

58

suspension concentration, volume and viscosity. The gap clearance is commonly used as main

59

operational parameter due to its fast and precise changeability.28,29 Lahtinen et al.30 used instead

60

the total operating power as primary control parameter. Due to the thermal expansion of the

61

grinding discs during processing, the total operating power proved to be a more reliable operating

62

parameter than the gap clearance. However, the total operating power strongly depends on the

63

grinder model (i.e. on the motor size and rotational speed of the grinding discs). Using the

64

operating power surplus, namely the difference between the total and the no-load power, as main

65

parameter to calculate the specific energy consumption is thus the most representative, yet non-

66

standardized way to describe and operate the grinding process.28 Commonly, low specific energy

67

consumptions between 1.5 and 5 kWh kg–1 were found to be sufficient for producing CNF films

68

with the most improved mechanical properties.17,31,32

69

The absence of a common quality index for CNF suspensions and materials restricts, however,

70

the comparison between the produced CNFs. There are different approaches to assess the degree

71

of fibrillation of CNFs in the suspension state, such as determining the degree of polymerization

72

of cellulose chains or calculating the water retention value.33,34 The only approach to a universal

73

quality index was recently proposed by Desmaisons et al.32 They selected eight widely available,

74

simple, and time-efficient criteria to characterize CNF suspensions produced using a grinder at

75

different total energy consumptions. The focus was on the suspension properties such as the macro-


4

Page 5 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


76

and nanosized fractions, turbidity, and homogeneity, but also on the indirect assessment of the

77

structural and mechanical properties of the corresponding CNF films. The indirect evaluation of

78

the CNF quality via film preparation gives additional information on e.g. CNF interactions and

79

specific surface area.32 This approach can be argued, as no common protocol for the preparation

80

and characterization of CNF films exists, leading to deviations in film properties and consequently

81

to differences in resolved CNF quality.35 In comparison to the various methods for producing CNF

82

films, directional ice templating (or freeze-casting) of nanocellulose suspensions is a reproducible

83

and straightforward technique to produce anisotropic CNF foams.10,36 Processing parameters such

84

as the freezing rate and direction during freezing of the CNF suspensions can be easily controlled,

85

allowing for reliable and reproducible evaluation of the CNF quality.37 CNF foams are, besides,

86

getting more and more attention, for a wide range of applications such as biomedical scaffolds,

87

thermal insulation or energy storage devices,38,39 as they combine light weight, tunable porous

88

structure, low thermal conductivity, and high compressive strength.38 Assessing and optimizing

89

the degree of fibrillation of CNFs is key for tailoring the properties of CNF foams for a target

90

application.

91

Here, we investigate the properties of anisotropic CNF foams obtained by directional ice

92

templating to evaluate the quality of CNFs. We propose a protocol allowing for an energy- and

93

time-efficient production of CNFs, which considers the specific energy consumption as unique

94

driving setting. The compressive strength, the porous morphology and CNF alignment in the cell

95

walls of anisotropic freeze-cast CNF foams were quantified and correlated to the CNF fibrillation

96

quality by combining image analysis algorithms, anisotropic thermal conductivity measurements,

97

and X-ray diffraction analysis.

98


5


99

Page 6 of 37

EXPERIMENTAL SECTION

100

Materials. Never-dried bleached Sulfite and Kraft softwood pulps were kindly provided by

101

Domsjö Fabriker AB (Sweden) and SCA Östrand (Sweden), respectively, and used as starting

102

materials for the CNF production. The pulps were treated with 0.5 M hydrochloric acid, ion-

103

exchanged to sodium form using 10-3 M sodium bicarbonate and washed thoroughly until neutral

104

pH with deionized water. The cellulose, hemicelluloses and lignin contents of the washed pulps

105

were determined by MoRe Research (Örnsköldsvik, Sweden) and are presented in Table S1, in the

106

Electronic Supplementary Information (ESI) available.

107 108

Preparation of the CNF suspensions. Washed pulp fibers were suspended in deionized water

109

at initial solid content of 2 wt% and dispersed for 5 min using a High-Shear Dispermix (Ystral

110

GmbH, Germany). The suspensions were passed through a supermasscolloider grinder (Model

111

MKZA10-15J, Masuko Sangyo Co. Ltd., Japan), equipped with non-porous grinding stones

112

containing silicon carbide (Disk model MKE 10-46#), for mechanical fibrillation. The rotational

113

speed of the grinding stones was set to 750 rpm (25 Hz motor frequency) and the zero contact

114

position between the two grinding stones was determined right before loading the fiber suspension.

115

Immediately after feeding the fiber suspension into the grinder, the gap clearance between the

116

grinding stones was set so that a certain surplus in operating power was reached. The no-load

117

power was 3.4 kW and the operating power during CNF production was held constant at 3.6 kW

118

by adapting the gap clearance between -70 and -200 μm. The suspensions were circulated through

119

the grinder and samples were taken according to prior-calculated energy consumption values.

120


6

Page 7 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


121

Energy consumption determination. The energy consumption during grinding was determined

122

based on the direct measurement of the operating power using a built-in three-phase wattmeter,

123

and the time of fibrillation. The specific energy consumption Es was calculated as followed:

124

𝐸𝑠 [𝑘𝑊ℎ 𝑘𝑔−1 ] =

𝑃𝑠 [𝑘𝑊]×𝑡[ℎ] 𝑤𝐶𝑁𝐹 [𝑘𝑔]

(1)

125

where Ps is the operating power surplus, which was determined by subtracting the no-load

126

operating power (3.4 kW) from the total operating power during grinding, t is the sampling time

127

and wCNF is the weight of dry CNFs processed for certain time t.

128 129

Characterization of the CNF suspensions. The morphology of the CNFs was observed by

130

scanning electron microscopy (SEM, JEOL JSM-7401F, JEOL Ltd., Japan). A drop of 0.001 wt%

131

CNF suspension was spread on a glass disc mounted onto a metal substrate using carbon tape,

132

allowed to dry overnight at room temperature and coated with a thin layer of gold. Images were

133

recorded at an accelerating voltage of 2 kV and a working distance of 8 mm. A minimum of 5

134

images were taken for each sample at different locations to obtain a representative overview of the

135

suspension. Diameter measurements of fibers were performed on high-magnification SEM images

136

using the ImageJ software. The aspect ratio, A, of the CNFs was determined by sedimentation

137

using the crowding number theory, as previously reported by Varanasi et al.40 and Martinez et al.41

138

Details on the procedure and calculations can be found in the ESI. Sedimentation experiments

139

were conducted twice per CNF suspension to ensure reproducibility.

140 141

Production of CNF foams by directional ice templating. 5 g of CNF suspension at 0.5 wt%

142

were poured into a cylindrical Teflon mold equipped with a copper bottom plate. The bottom of

143

the mold was put in contact with dry ice, until complete freezing of the CNF suspension. The


7


Page 8 of 37

144

cooling rate was measured to be around 3 K min–1. The frozen CNF suspension was then

145

sublimated using a freeze-dryer (Alpha 1-2 LDplus, Christ, Germany) for at least 72 h at a pressure

146

of 0.0024 mbar. 20 foams were prepared for each CNF grade.

147 148

Characterization of the CNF foams. The apparent density (ρapp) was calculated from the mass

149

and volume of the foams after conditioning for at least 48 h at 23 °C and 50% relative humidity

150

(RH). The porosity (ε) was calculated as (1–ρrel) × 100 % where ρrel is the relative density. The

151

relative density is calculated as ρapp/ρskel where ρskel is the skeletal density of cellulose (1500 kg m-

152

3 42

153

Instrument Corporation, Nocross, GA, USA). The BET (Brunauer–Emmet–Teller) model was

154

used to estimate the surface area of the foams. The CNF foams were degassed at 80 °C for 10 h

155

prior to the measurements. The porous structure in the axial and radial directions (i.e. parallel and

156

perpendicular to the ice crystal growth) of the CNF foams was assessed on specimens coated with

157

a thin layer of gold using a TM 3000 SEM (Hitachi High Tech, Japan) at 5 kV accelerating voltage.

158

The pore sizes were estimated using the ImageJ software.

).

Nitrogen sorption measurements were performed using an ASAP 2020 (Micromeritics

159

The local orientation degree of the CNFs in the foam axial direction was determined by SEM

160

image analysis using the ImageJ plug-in “OrientationJ”.43 The image analysis calculates the

161

structure tensor of each pixel of coordinates (x; y) in its local neighborhood. The local orientation

162

and isotropy, i.e. coherency and energy, respectively, of every pixel in the image is evaluated by

163

computing the spatial derivatives in x and y directions using a cubic spline gradient.44 Histograms

164

of the local orientation weighed by the coherency were obtained for each image and they were

165

fitted to a Gaussian function for assessing the orientation index, f. f was calculated using the full

166

width at half-maximum (fwhm) of the Gaussians as follows:45


8

Page 9 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


167 168 169

𝑓=

180−𝑓𝑤ℎ𝑚 180

(2)

Analyses were conducted on at least ten SEM images of three different foam pieces along the axial direction.

170

The compressive behavior of the foams in the axial direction was evaluated using an Instron

171

5944 mechanical testing instrument (Instron, USA) equipped with a 500 N load cell. The foams

172

were conditioned for at least 48 h at 23 °C and 50% relative humidity (RH) prior to the

173

measurements. The mechanical testing was conducted at 23 °C and 50% RH at a strain rate of 10%

174

min−1. The compressive Young’s modulus was determined from the slope of the initial linear

175

region of the stress−strain curve and the energy absorbed by the foam (i.e. toughness) was

176

evaluated from the area under the stress−strain curve up to 70% strain. The average value and

177

standard deviations of measurements on three to four foam specimen are reported.

178

X-ray diffraction (Oxford Diffraction Xcalibur 3, Agilent Technologies, USA) was used to

179

characterize the Hermans’ orientation parameter of the CNFs in the foam cell walls. Cylindrical

180

foams were compressed parallel to the axial direction and 2D patterns were recorded with a

181

sample-to-detector distance of 45 mm. The beam source was molybdenum (Mo Kα1, λ = 0.71 nm),

182

the exposure time was set at 2 × 300 s, and the CCD detector used was a Sapphire 3. The Hermans’

183

orientation parameter, fH, quantitatively describes the alignment of the CNFs relative to the

184

freezing direction (i.e. axial plane).45,46 fH is obtained by azimuthal integration of the (200) peak

185

of cellulose (θ = 11.4°) and the following equations:46

186 187

𝑓𝐻 = 〈𝑐𝑜𝑠 2 𝜑〉 =

3〈𝑐𝑜𝑠 2 𝜑〉−1 2

2 ∑𝜋/2 0 𝐼(𝜑) 𝑠𝑖𝑛 𝜑𝑐𝑜𝑠 𝜑 𝜋/2

∑0

𝐼(𝜑) 𝑠𝑖𝑛 𝜑

(3) (4)


9


Page 10 of 37

188

where φ represents a theoretical angle between the main direction of the nanofibril and the

189

freezing direction of the ice crystals during ice templating. This angle can be identified as the

190

azimuthal angle on the 2D pattern. I(φ) represents the intensity at a certain φ angle.

191

The axial and radial thermal conductivities (λ, mW/mK) of the anisotropic CNF foams were

192

measured in anisotropic mode using the TPS 2500 S Thermal Constants Analyzer (Hot Disk,

193

Sweden).47 The heating power was set at 20 mW and the measurement time was 10 sec for each

194

thermal conductivity measurement. The foams were enclosed in a customized cell, allowing the

195

relative humidity (RH) to be controlled at 50 % RH using a P2 humidifier (Cellkraft, Sweden).

196

The temperature of the foams was controlled (295 K) by immersing the customized cell in a

197

temperature controlled silicon oil bath. Five independent measurements were performed with 15

198

min interval time on two pairs of the investigated foams. The foams were kept at the set

199

temperature and RH for at least 120 min prior to measurements of the thermal conductivity. The

200

foams density and specific heat capacity at 50 % RH were given as input for performing the

201

measurements in anisotropic mode (see ESI). The thermal conductivity anisotropy ratio (AR) was

202

calculated by dividing the axial thermal conductivity values by the radial ones.

203 204

RESULTS AND DISCUSSION

205

Mechanical fibrillation via grinding and energy consumption

206

The Kraft and Sulfite CNF suspensions were sampled at defined specific energy consumptions

207

during grinding. We selected much lower energy consumption values than the ones reported in the

208

literature to investigate the influence of (ultra)low energy consumption on the fibrillation

209

quality.17,31,32,48,49 The first three CNF suspensions, namely K1, K2 and K3 from Kraft pulp and

210

S1, S2 and S3 from Sulfite pulp, were obtained after one, two and three passes through the grinder.


10

Page 11 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


211

The amount of passes corresponds to specific energy consumptions of 0.12, 0.44, 0.92 kWh kg–1

212

and 0.12, 0.39, 0.85 kWh kg–1, respectively. In addition, two CNF specimen were sampled after 3

213

kWh kg–1 (K4, S4) and 5 kWh kg–1 (K5, S5) for comparison with CNF suspensions obtained at

214

higher energy consumption,17 but of homogeneous and good quality, as reported by Desmaisons

215

et al.32 The number of passes that the suspension went through the grinder at each selected specific

216

energy consumption is depicted in Fig. 1. For the production of Sulfite CNFs the energy

217

consumption increased with a constant rate of 0.36 kWh kg–1 for each pass (Linear regression

218

performed with R2 = 0.99). The Kraft pulp, however, required a longer grinding time per pass

219

compared with the Sulfite pulp, which resulted in an increase in specific energy consumption for

220

each grinding cycle (Fig. 1). These results are in accordance with prior findings, showing that the

221

presence of hemicelluloses in Kraft pulp (Table S1 in the ESI) induces a better fibrillation and an

222

increase in viscosity during grinding. Hemicelluloses act as inhibitors of the coalescence of

223

fibrils.50,51 For comparison, Fig. 1 also reports the energy consumption and number of passes used

224

by Josset et al.31 for producing CNFs from bleached Kraft softwood pulp containing 12.6 wt%

225

hemicelluloses (against 16.2 wt% and 3.8 wt% for the Kraft and Sulfite pulps in the present case,

226

Fig. 1 inset). They reported a linear behavior between the number of passes and energy

227

consumption, while here this correlation deviated towards higher energy input per pass for Kraft

228

CNFs. This suggests that from a certain hemicelluloses content, the suspension viscosity increases

229

due to promoted fibrillation, which results in a non-linear behavior between the number of passes

230

and the specific energy consumption of the system.


11


Page 12 of 37

231 232

Figure 1. Correlation between the number of passes through the grinder and the specific energy

233

consumption applied during grinding of the Kraft (orange ) and Sulfite (blue ○) pulps for CNF

234

production. Data from the study of Josset et al.26 (grey ) is included for comparison. Inset graph:

235

Comparison of hemicelluloses content in the three distinct pulps.

236 237

Morphology and aspect ratio of CNFs

238

The morphology and size distribution of the CNFs were qualitatively evaluated by SEM

239

imaging. Optimal visualization of the nanofibril quality using microscopy techniques remains

240

challenging as the whole fibrillation state cannot be accurately assessed due to e.g. exclusion of

241

large fibers at high magnification, or partial analysis of the CNF batches produced.52 Due to the

242

tight CNF entanglement, the width and length of the different materials can, besides, not be

243

distinctively measured.53 An overview of the fibrillation state at the different specific energy

244

consumptions used to produce the CNFs is shown in Fig. 2.


12

Page 13 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


245 246

Figure 2. SEM images of CNF suspensions obtained after grinding with different specific energy

247

input. Top row: Sulfite pulp CNFs after 0.12 kWh kg–1 (S1), 0.85 kWh kg–1 (S3) and 5 kWh kg–1

248

(S5). Bottom row: Kraft pulp CNFs after 0.12 kWh kg–1 (K1), 0.92 kWh kg–1 (K3) and 5 kWh kg–

249

1

(K5). The images were taken at same magnification (×2000). The scale bar is 10 μm.

250 251

The first low energy input selected (corresponding to one pass through the grinder) was

252

sufficient to start fibrillating both the Kraft and Sulfite pulps as shown on Fig. 2 (K1 and S1). At

253

very low specific energy consumptions (< 1 kWh kg–1), the Sulfite CNF suspensions still presented

254

microfibril aggregates with diameters up to several micrometers and some short intact fibers (S1-

255

3; S2 in Fig. S1 in the ESI). Higher energy input induced a more homogeneous fibrillation of the

256

Sulfite pulp (S4-5; S4 in Fig. S1 in the ESI). The fibers were disintegrated into CNFs (diameters

257

measured below 100 nm using high-magnification SEM images as shown in Figure S2 in the ESI)

258

and microfibrils up to around one micrometer in width. With a higher hemicelluloses content, the

259

Kraft pulp disintegrated right away at early grinding stages compared with the Sulfite pulp. The

260

presence of hemicelluloses in the pulp was reported to regulate nanofibril aggregation by hydrogen


13


Page 14 of 37

261

bonding and electrostatic repulsions between hemicelluloses’ carboxylates, thus promoting the cell

262

wall delamination.32,50,51 The Kraft pulp fibers were fibrillated down to the nanofibril level (Figure

263

S2 in the ESI) with almost no residual intact fibers left, after one grinding cycle only (K1). With

264

increasing energy input, disintegration of the pulp fibers towards nanofibrils got more

265

homogeneous with sporadic presence of microfibrils (K2-K5, K2 and K4 in Fig. S1 in the ESI).

266

Assessment of the average CNF aspect ratio was carried out by sedimentation of the Kraft and

267

Sulfite pulps and respective CNFs (Fig. 3).40,41 Right after one grinding cycle, we observed a steep

268

increase in the aspect ratio of both the Sulfite and Kraft fibrillated materials (K1 and S1). This

269

increase may result from a combination of transversal and longitudinal fibrillation of the pulp

270

fibers when exposed to high-shear forces between grinding stones. Reduction in the fiber diameters

271

is more probable to occur during the first grinding cycle, as it results in a drastic increase of the

272

fibers’ aspect ratio. SEM images confirm this trend showing microfibrils with much lower widths

273

than that of the pulp fibers (K1-3 and S1-3). Additional specific energy input led to a constant

274

average aspect ratio (K2-3 and S2-3). From 3 kWh kg–1, however, a gradual decrease in aspect

275

ratio is observed (K4-5 and S4-5) due to further transversal fibrillation. The changes in aspect ratio

276

are more pronounced for the Kraft CNFs than the Sulfite CNFs: the aspect ratio increased by +123

277

% between K0 (73) and K3 (162) against +40 % for S0 (73) and S3 (102), respectively. This

278

increase was followed up by a 39 % decrease to 98 (K5) against a 33 % decrease to 68 (S5). These

279

results support the previous microscopic analysis and confirm that a higher content of

280

hemicelluloses in the pulp promotes the fibrillation of the fibers to CNFs. Compared with Sulfite

281

pulp, Kraft pulp clearly led to more energy-efficient and homogeneous fibrillated materials. This

282

pulp has thus been selected for assessing further the fibrillation quality via the investigation of

283

anisotropic CNF foams’ properties.


14

Page 15 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


284 285

Figure 3. Evolution of the aspect ratio (L/D) of the Sulfite (blue ○) and Kraft (orange ) pulps

286

and CNFs determined by sedimentation, as a function of the specific energy consumption during

287

grinding.

288 289

Porous architecture of the CNF foams

290

Directional ice templating was carried out for producing anisotropic CNF foams, whose cell

291

walls consist of aligned CNFs.54 The average density of the foams was 5.5 ± 0.20 kg m–3 which

292

resulted in a porosity of 99.63 ± 0.01 %. The BET specific surface areas were in the range of 22 –

293

29 m2 g-1 for all foams. Fig. 4 shows SEM images of the anisotropic CNF foams cross-sections.


15


Page 16 of 37

294 295

Figure 4. SEM images of the Kraft pulp (K0) and Kraft CNF (K1-5) foams’ cross-section. K1 to

296

K5 CNFs were produced using a grinder at different specific energy consumption, from 0.12 kWh

297

kg–1 (K1) to 5 kWh kg–1 (K5), respectively. For each image, the magnification is ×100 and the

298

scale bar is 500 μm.

299 300

Foams prepared from the Kraft pulp showed no defined cell shape (Fig. 4, K0). According to

301

Wegst et al.,55 two main factors determine whether freeze-cast particles are trapped inside the ice

302

crystals or rejected and pushed aside, namely (i) the freezing front velocity and (ii) the balance

303

between attractive and repulsive forces acting on particles. While the freezing front velocity was

304

determined to be almost identical for all CNF suspensions (ca. 1.5 mm min–1), the difference in

305

particle engulfment should arise from the differences in forces acting on the particle. The repulsive

306

force, FR, that a spherical particle with a radius r experiences in the liquid phase is caused by

307

molecular van der Waals interactions at the solid-liquid interface and can be expressed according

308

to Equation 5:55


16

Page 17 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


𝑎

𝐹𝑅 = 2𝜋𝑟𝛥𝜎0 ( 0)

309

𝑑

(5)

310

where 𝛥𝜎0 is the free energy of the system, d is the thickness of the liquid layer between the

311

particle and the freezing front and 𝑎0 is the mean distance between molecules in this liquid layer.55

312

The attractive force FA working on the same particle is generated by the viscous drag of the moving

313

liquid and derived from the Stokes law according to Equation 6:55

314

𝐹𝐴 =

6𝜋𝜂𝜈𝑟 2 𝑑

(6)

315

where η is the dynamic viscosity of the liquid and ν is the freezing front velocity. After balancing

316

those two forces and considering all parameters, except the particles’ radius, constant, the repulsive

317

and attractive forces scale both with the particles’ radius. Repulsive forces directly scale with the

318

radius, whereas attractive forces scale with the radius squared. Thus, bigger particles will be more

319

likely attracted to the solid-liquid interface at the freezing front and consequently entrapped by the

320

growing ice crystals. When extrapolating this theory to non-spherical particles such as fibers, the

321

entangled fiber network observed on Fig. 4-K0 therefore results from the large dimensions of the

322

cellulose fibers. By increasing the specific energy consumption, the particle length and diameter

323

were reduced (Fig. 3), resulting in a decrease of the attractive forces between the particle (here,

324

CNFs) and the freezing front, and in an increase in the probability for the CNFs to be rejected from

325

the solid phase. Foams made from K1 to K5 CNF suspensions present a more defined porous

326

structure than the K0 foam with pore sizes bigger than 30 μm, but the presence of residual

327

macrofibrils and poorly fibrillated material in K1-3 suspensions generated defective cell

328

boundaries and non-homogeneous cell size distributions (Fig. 4, K1 to K3). With a minimum

329

specific energy consumption of 3 kWh kg–1, the foam porous structure became significantly more

330

homogeneous, and formed a more regular and defined open honeycomb-like structure with less

331

defective cell walls (Fig. 4, K4 and K5).


17


Page 18 of 37

332 333

Compressive behavior of the CNF foams

334

The CNF foams were compressed in the strong and stiff axial direction, i.e. in the out-of-plane

335

deformation of the honeycombs, behaving as a cellular solid during compressive deformation.42

336

The compression started with an initial cell wall bending, which is related to a linear elastic

337

behavior and defined by an apparent Young’s modulus for early stress and strain. With further

338

compression, the collapse of the cells followed, which was shown by a plateau region of plastic

339

deformation. For a strain range from 70 to 90 %, a drastic stress increase was observed, indicating

340

the beginning of the densification regime, i.e. when the cell walls get in contact with each other.42

341

Fig. 5 compares the compressive Young’s modulus and energy absorption of the different CNF

342

foams.

343 344

Figure 5. Influence of the specific energy consumption on the energy absorption (orange ) and

345

the compressive Young’s modulus (blue ○) of anisotropic CNF foams (Axial compression parallel

346

to the freezing direction, i.e. CNF alignment).

347


18

Page 19 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


348

Foams made from the initial pulp fibers suspension (K0) had a very low energy absorption and

349

almost no detectable Young’s modulus resulting from the lack of fibers alignment. Owing to the

350

honeycomb-like porous structure of the K1, K2 and K3 foams (Fig. 4), a significant increase in

351

both the energy absorption (ca. 8-fold, K3) and Young’s modulus (ca. 26-fold, K3) was observed.

352

The compressive properties of the foam improved globally with the specific energy consumption

353

up to 3 kWh kg–1. As the amount of CNFs and thus, the overall share of fibers with small diameters,

354

increased with the specific energy input, more and more fibril-to-fibril bonds were formed,

355

resulting in an entangled network that is more efficient for stress transfer and distribution.24 The

356

improvement in fibril interactions due to the H-bonding capacity of residual hemicelluloses also

357

promoted the formation of a more cohesive material that better withstands external forces during

358

compression.51 From 3 kWh kg–1, however, the energy absorption only increased by 7.5 % (from

359

K4 to K5), while the increase in energy input amounted to 67 %. In the case of the Young’s

360

modulus, the plateau from 3 kWh kg–1 is concluded as well, as K4 and K5 foams showed similar

361

Young’s moduli, confirming the optimum value to reach for combining an energy-efficient

362

grinding with optimal CNF foam mechanical properties.

363 364

Characterization and quantification of the CNFs alignment

365

SEM images of the foams’ longitudinal section revealed the expected oriented porous structure

366

formed by the alignment of CNFs during freezing. The plugin OrientationJ was used to quantify

367

the degree of local orientation observed on the SEM images. Fig. 6 shows the different steps of

368

the image analysis. Representative SEM images of K1 and K4 foams (Fig. 6a and b, respectively)

369

were selected for clear distinction between two different fibrillation states and orientation degrees

370

of CNFs.


19


Page 20 of 37

371 372

Figure 6. Quantification of the CNF alignment in the foams’ cell walls by image analysis: (a,b)

373

SEM images of the foams K1 and K4 in the axial direction, respectively, (c,d) their respective

374

color-coded HSB maps (Hue: Orientation, Saturation: Coherency, Brightness: Original image),

375

(e,f) the corresponding OrientationJ output histograms (black curve) including respective Gaussian

376

fitting (red curve), and (g) influence of the specific energy consumption on the orientation index

377

calculated using Equation 2 for each anisotropic CNF foam.

378 379

From the original SEM images, color-coded maps were created by the software, taking into

380

consideration the original image brightness, the coherency as saturation and the orientation as hue

381

setting. A higher number of red pixels illustrates a better alignment of the CNFs in the freezing

382

direction, 0°, (horizontal direction in Fig. 6), as shown Fig. 6d. A greater share of green and blue

383

color tones as observed in Fig. 6c indicates CNF orientation deviating from the freezing direction.

384

The histograms of local orientation for K1 (Fig. 6e) and K4 (Fig. 6f) show a peak of different

385

width centered at 0°. The computed histograms (black curves) were fitted with a Gaussian function

386

(red curves) to calculate the orientation index, f, using Equation 2 (Fig. 6g). The foams produced

387

with low-energy consumption CNF suspensions (K1-3) exhibited less homogeneous and oriented


20

Page 21 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


388

cell walls than the foams produced with the K4 and K5 suspensions (Fig. 4 and Fig. 6a,b). The K1,

389

K2, and K3 foams show, indeed, a lower local orientation index with much greater standard

390

deviations than the K4 and K5 foams (Fig. 6g), which supports the SEM analysis of the foams’

391

cross-sections and the trend in mechanical properties. A higher local orientation degree is thus

392

assessed from 3 kWh kg–1 in specific energy consumption, representing the minimal value to reach

393

for achieving the highest CNF alignment (K4 and K5 foams).

394

The structural anisotropy of the foams can also be highlighted at a finer scale than that of the

395

previous macroscopic attempts. One commonly reported method to estimate the alignment of rod-

396

like particles is X-ray diffraction.36,45 During freeze-casting, the shear forces induced by the ice

397

crystals growth and the resulting confinement cause the fibrils to pack, to a certain extent, parallel

398

to the freezing direction. Quantifying the orientation of the cellulose crystal units along the

399

freezing direction using X-ray diffraction may be another tool to assess the fibrillation quality of

400

the CNF suspensions. The Hermans’ orientation parameter, fH, quantifies the crystals alignment

401

(Equation 3). The closer fH is to 1, the more the crystal units of the CNFs are aligned. fH was

402

calculated for the foams K0 to K5 (Fig. S3 in the ESI). K0 displays a value of fH = 0.01, which in

403

theory corresponds to a completely random distribution,46 meaning that the shear forces were

404

insufficient to induce any orientation during freezing. All the other CNF foams, in comparison,

405

have fH values ranging between 0.20 and 0.31, indicating a moderate alignment. To put these

406

values into perspective, they are within the same range as values obtained for drawn films of

407

regenerated cellulose (up to fH = 0.29),56 and are much lower than the coefficients calculated for

408

drawn films made of TEMPO-oxidized CNFs (fH = 0.34 to 0.72),45 which is to be expected given

409

the particles’ sizes. It is however difficult to state that fH allows for an effective discrimination in

410

the fibrillation degree, considering the similarity between the values obtained for K3 and K5 and


21


Page 22 of 37

411

the low value of K4. Because the mechanically treated CNF suspensions did not show any

412

significant differences in fH values, XRD is in the present case not the optimal experimental tool

413

for assessing the fibrillation quality. However, when considering previous studies reporting clear

414

differences in fH values between e.g. TEMPO-CNFs and regenerated cellulose,45,56 determining

415

the fH orientation parameter can be a satisfactory tool to estimate the fiber orientation degree at the

416

single fibril level. The use of fH seems more appropriate for comparing different systems (e.g.

417

different types of CNFs or processing) than for tracking the evolution of a single system over the

418

course of one processing method. In the present case, assessing the macroscopic properties of the

419

foams is more reliable to evaluate the fibrillation quality of the CNFs.

420

The foams’ thermal conductivity was evaluated as another indirect method to quantify the CNF

421

alignment in the foams in correlation with the fibrillation quality. The thermal conductivity at room

422

temperature of CNF foams with cell sizes below 1 mm is dependent on two main contributions, (i)

423

the solid contribution related to the heat conduction in the cell walls, and (ii) the gas contribution

424

corresponding to the conduction of air in the macropores.57–59 Evaluation of the anisotropy of the

425

CNF foams was conducted via the measurement of the thermal conductivity in axial and radial

426

directions. Fig. 7a shows the axial and radial thermal conductivity values of the different CNF

427

foams determined at 50 % relative humidity and 22 °C.


22

Page 23 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


428 429

Figure 7. Thermal conductivity of anisotropic CNF foams: an indirect quantitative method of the

430

CNF alignment. (a) Axial (blue) and radial (orange) thermal conductivity of the foams measured

431

at 50 % RH and 22 °C. (b) Anisotropy ratio of thermal conductivity defined as the ratio of the axial

432

by the radial thermal conductivity values. Anisotropy values greater than 1 indicate CNF alignment

433

in the axial direction.

434 435

As previously reported,10 the axial thermal conductivity of anisotropic CNF foams is

436

significantly higher than the radial one due to an important solid contribution of the cell walls in

437

the axial direction. Previous studies have indeed shown that the gas conduction inside macroporous


23


Page 24 of 37

438

solids is very close to the gaseous conduction in free space (25 mW m–1K–1) and cannot account

439

for the significantly higher axial thermal conductivity.60 The axial thermal conductivity increased

440

with the increasing specific energy input from 119 to 132 mW m–1K–1 (K1 to K3) and levelled off

441

around 147 ± 5 mW m–1K–1 from 3 kWh kg–1 (K4-5). The opposite trend was observed for the

442

radial thermal conductivity. It is worth noting that each CNF foam behaves as a super-insulating

443

material in the radial direction (i.e. thermal conductivity, λ < λair = 25 mW m–1K–1).10 When

444

calculating the anisotropy ratio (Fig. 7b), an increase in the anisotropy is observed up to 3 kWh

445

kg–1. The CNFs obtained with a higher specific energy input resulted in more anisotropic

446

structures, which can be compared to the structure of TEMPO-oxidized CNF (T-CNF) foams made

447

by freeze-casting.10,36 Wicklein et al.10 reported an axial and radial thermal conductivity of 150

448

and 18 mW m–1 K–1, respectively, for T-CNF foams at 50 % RH and 22 °C. These values are

449

similar to those measured for the foam K4 made from the CNF suspension produced with a specific

450

energy consumption of 3 kWh kg–1. T-CNFs are produced from a combination of chemical and

451

mechanical treatments, which usually results in the most fibrillated CNF materials, reaching a

452

nanofibril width down to 3-4 nm.20 This level of fibrillation has so far never been achieved by

453

mechanical disintegration only. Yet, in this study, we obtained CNF foams exhibiting very similar

454

thermal conductivities to that of T-CNF foams, by solely using a mechanical fibrillation and

455

without achieving CNFs of 3-4 nm in width.

456

Fig. 8 summarizes the correlations observed between the CNF alignment, the mechanical and

457

thermal properties of the Kraft CNF foams as a function of the specific energy consumption used

458

to produce the CNFs. Each properties optimum converges to a specific energy consumption of 3

459

kWh kg–1, which is the optimum value to reach for producing the best CNF fibrillation quality by

460

energy-efficient grinding.


24

Page 25 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


461 462

Figure 8. Correlation between the degree of fibrillation set by the specific energy consumption

463

during grinding and the properties of the anisotropic CNF foams, namely the CNF orientation

464

index (grey ), the compressive Young's modulus (blue ○) and the anisotropy ratio of thermal

465

conductivity (orange ).

466 467

CONCLUSION

468

Kraft and Sulfite pulps were successfully fibrillated using the operating energy input surplus as

469

the main processing parameter during grinding. The presence of residual hemicelluloses in the

470

Kraft pulp promoted the energy-efficient disintegration of the cellulose fibers into CNFs and led

471

to the production of a more homogeneous material in size. The characterization of the CNF aspect

472

ratio by sedimentation resulted in a first overview of the degree of nanofibrillation. The analysis

473

of the anisotropic properties of the Kraft CNF foams prepared by directional ice templating was

474

presented as a novel approach to assess quantitatively the fibrillation quality of the CNFs. The

475

macroporous morphology and the compressive behavior of the CNF foams revealed a uniform

476

porous honeycomb structure and optimal compressive Young’s modulus and toughness at an

477

energy consumption of 3 kWh kg–1. The study of the CNF alignment in the foam cell walls via

478

novel techniques such as image analysis and thermal conductivity measurement confirmed the


25


Page 26 of 37

479

optimal plateau at 3 kWh kg–1. Correlating the properties of anisotropic CNF foams with the degree

480

of fibrillation of the CNFs is a reproducible, direct and promising approach, which paves the way

481

for energy-efficient production of CNFs without compromising the quality of the suspension.

482


26

Page 27 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


483

ASSOCIATED CONTENT

484

Supporting Information

485

The Supporting Information including additional information on materials and methods as well

486

as SEM images of the CNFs and quantification of CNF alignment is available free of charge on

487

the ACS Publications website at DOI: ______.

488

AUTHOR INFORMATION

489

Corresponding Author

490

*E-mail: [email protected] (Nathalie Lavoine)

491

ORCID

492

Konstantin Kriechbaum: 0000-0002-3737-5303

493

Nathalie Lavoine: 0000-0002-8259-4070

494

Present Address

495

‡

496

University, 2820 Faucette Dr., Raleigh, NC 27695, USA

497

Author Contributions

498

The manuscript was written through contributions of all authors. All authors have given approval

499

to the final version of the manuscript.

500

Notes

501

The authors declare no conflict of interest.

Department of Forest Biomaterials, College of Natural Resources, North Carolina State

502


27


Page 28 of 37

503

ACKNOWLEDGMENT

504

The authors acknowledge the Swedish Research Council for Environment, Agricultural Sciences

505

and Spatial Planning, the Swedish Energy Agency (Energimyndigheten), the Swedish Foundation

506

for strategic research (SSF) and the Marie Sklodowska-Curie Actions Innovative Training

507

Networks (H2020-MSCA-ITN-2014) for financial support. The authors would especially like to

508

thank Prof. Lennart Bergström for his support and Prof. Aji Mathew for her fruitful advice.

509 510

REFERENCES

511

(1)

Mariano, M.; El Kissi, N.; Dufresne, A. Cellulose Nanocrystals and Related

512

Nanocomposites: Review of Some Properties and Challenges. J. Polym. Sci. Part B Polym.

513

Phys. 2014, 52 (12), 791–806, DOI 10.1002/polb.23490.

514

(2)

Abitbol, T.; Rivkin, A.; Cao, Y.; Nevo, Y.; Abraham, E.; Ben-Shalom, T.; Lapidot, S.;

515

Shoseyov, O. Nanocellulose, a Tiny Fiber with Huge Applications. Curr. Opin. Biotechnol.

516

2016, 39, 76–88, DOI 10.1016/j.copbio.2016.01.002.

517

(3)

518 519

2008, 86 (6), 484–494, DOI 10.1139/v07-152. (4)

520 521

Dufresne, A. Polysaccharide Nano Crystal Reinforced Nanocomposites. Can. J. Chem.

Siró, I.; Plackett, D. Microfibrillated Cellulose and New Nanocomposite Materials: A Review. Cellulose 2010, 17 (3), 459–494, DOI 10.1007/s10570-010-9405-y.

(5)

Klemm, D.; Kramer, F.; Moritz, S.; Lindström, T.; Ankerfors, M.; Gray, D.; Dorris, A.

522

Nanocelluloses: A New Family of Nature-Based Materials. Angew. Chemie Int. Ed. 2011,

523

50 (24), 5438–5466, DOI 10.1002/anie.201001273.


28

Page 29 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

524


(6)

Siqueira, G.; Bras, J.; Dufresne, A. Cellulosic Bionanocomposites: A Review of

525

Preparation, Properties and Applications. Polymers (Basel). 2010, 2 (4), 728–765, DOI

526

10.3390/polym2040728.

527

(7)

528 529

Li, F.; Mascheroni, E.; Piergiovanni, L. The Potential of Nanocellulose in the Packaging Field: A Review. Packag. Technol. Sci. 2015, 28 (6), 475–508, DOI 10.1002/pts.2121.

(8)

Serpa, A.; Velásquez-Cock, J.; Gañán, P.; Castro, C.; Vélez, L.; Zuluaga, R. Vegetable

530

Nanocellulose in Food Science: A Review. Food Hydrocoll. 2016, 57, 178–186, DOI

531

10.1016/j.foodhyd.2016.01.023.

532

(9)

Brodin, F. W.; Gregersen, O. W.; Syverud, K. Cellulose Nanofibrils: Challenges and

533

Possibilities as a Paper Additive or Coating Material - a Review. Nord Pulp Pap Res J 2014,

534

29 (1), 156–166, DOI 10.3183/NPPRJ-2014-29-01-p156-166.

535

(10) Wicklein, B.; Kocjan, A.; Salazar-Alvarez, G.; Carosio, F.; Camino, G.; Antonietti, M.;

536

Bergström, L. Thermally Insulating and Fire-Retardant Lightweight Anisotropic Foams

537

Based on Nanocellulose and Graphene Oxide. Nat. Nanotechnol. 2015, 10, 277–283, DOI

538

10.1038/nnano.2014.248.

539

(11) Kobayashi, Y.; Saito, T.; Isogai, A. Aerogels with 3D Ordered Nanofiber Skeletons of

540

Liquid‐Crystalline Nanocellulose Derivatives as Tough and Transparent Insulators. Angew.

541

Chemie Int. Ed. 2014, 53 (39), 10394–10397, DOI 10.1002/anie.201405123.

542

(12) Yildirim, N.; Shaler, S. M.; Gardner, D. J.; Rice, R.; Bousfield, D. W. Cellulose Nanofibril

543

(CNF) Reinforced Starch Insulating Foams. Cellulose 2014, 21 (6), 4337–4347, DOI

544

10.1007/s10570-014-0450-9.


29


545 546 547 548

Page 30 of 37

(13) Gatenholm, P.; Klemm, D. Bacterial Nanocellulose as a Renewable Material for Biomedical Applications. MRS Bull. 2010, 35 (3), 208–213, DOI 10.1557/mrs2010.653. (14) Jorfi, M.; Foster, E. J. Recent Advances in Nanocellulose for Biomedical Applications. J. Appl. Polym. Sci. 2015, 132 (14), DOI 10.1002/app.41719.

549

(15) Sultan, S.; Mathew, A. 3D Printed Scaffolds with Gradient Porosity Based on Cellulose

550

Nanocrystal Hydrogel. Nanoscale 2018, 10, 4421–4431, DOI 10.1039/C7NR08966J.

551

(16) Hoeng, F.; Denneulin, A.; Bras, J. Use of Nanocellulose in Printed Electronics: A Review.

552

Nanoscale 2016, 8 (27), 13131–13154, DOI 10.1039/C6NR03054H.

553

(17) Spence, K. L.; Venditti, R. A.; Rojas, O. J.; Habibi, Y.; Pawlak, J. J. A Comparative Study

554

of Energy Consumption and Physical Properties of Microfibrillated Cellulose Produced by

555

Different Processing Methods. Cellulose 2011, 18 (4), 1097–1111, DOI 10.1007/s10570-

556

011-9533-z.

557

(18) Henriksson, M.; Henriksson, G.; Berglund, L. A.; Lindström, T. An Environmentally

558

Friendly Method for Enzyme-Assisted Preparation of Microfibrillated Cellulose (MFC)

559

Nanofibers.

560

10.1016/j.eurpolymj.2007.05.038.

561 562

Eur.

Polym.

J.

2007,

43

(8),

3434–3441,

DOI

(19) Aulin, C. Novel Oil Resistant Cellulosic Materials, PhD thesis, KTH Royal Institute of Technology, 2009.

563

(20) Saito, T.; Kimura, S.; Nishiyama, Y.; Isogai, A. Cellulose Nanofibers Prepared by TEMPO-

564

Mediated Oxidation of Native Cellulose. Biomacromolecules 2007, 8 (8), 2485–2491, DOI

565

10.1021/bm0703970.


30

Page 31 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

566 567


(21) Li, Q.; McGinnis, S.; Sydnor, C.; Wong, A.; Renneckar, S. Nanocellulose Life Cycle Assessment. ACS Sustain. Chem. Eng. 2013, 1 (8), 919–928, DOI 10.1021/sc4000225.

568

(22) Arvidsson, R.; Nguyen, D.; Svanström, M. Life Cycle Assessment of Cellulose Nanofibrils

569

Production by Mechanical Treatment and Two Different Pretreatment Processes. Environ.

570

Sci. Technol. 2015, 49 (11), 6881–6890, DOI 10.1021/acs.est.5b00888.

571

(23) Li, Q.; Renneckar, S. Molecularly Thin Nanoparticles from Cellulose: Isolation of Sub-

572

Microfibrillar Structures. Cellulose 2009, 16 (6), 1025–1032, DOI 10.1007/s10570-009-

573

9329-6.

574

(24) Dufresne, A.; Cavaille, J.-Y.; Vignon, M. R. Mechanical Behavior of Sheets Prepared from

575

Sugar Beet Cellulose Microfibrils. J. Appl. Polym. Sci. 1997, 64 (6), 1185–1194, DOI

576

10.1002/(SICI)1097-4628(19970509)64:6%3C1185::AID-APP19%3E3.0.CO;2-V.

577

(25) Zhang, L.; Tsuzuki, T.; Wang, X. Preparation of Cellulose Nanofiber from Softwood Pulp

578 579 580

by Ball Milling. Cellulose 2015, 22 (3), 1729–1741, DOI 10.1007/s10570-015-0582-6. (26) Turbak, A. F.; Snyder, F. W.; Sandberg, K. R. Suspensions Containing Microfibrillated Cellulose, 1985, Patent US4500546A

581

(27) Iwamoto, S.; Nakagaito, A. N.; Yano, H. Nano-Fibrillation of Pulp Fibers for the Processing

582

of Transparent Nanocomposites. Appl. Phys. A Mater. Sci. Process. 2007, 89 (2), 461–466,

583

DOI 10.1007/s00339-007-4175-6.

584

(28) Kang, T.; Paulapuro, H. New Mechanical Treatment for Chemical Pulp. Proc. Inst. Mech.

585

Eng.

Part

E

J.

Process

586

10.1243/09544089JPME81.

Mech.

Eng.

2006,

220

(3),

161–166,

DOI


31


Page 32 of 37

587

(29) Hu, C.; Zhao, Y.; Li, K.; Zhu, J. Y.; Gleisner, R. Optimizing Cellulose Fibrillation for the

588

Production of Cellulose Nanofibrils by a Disk Grinder. Holzforschung 2015, 69 (8), 993–

589

1000, DOI 10.1515/hf-2014-0219.

590

(30) Lahtinen, P.; Liukkonen, S.; Pere, J.; Sneck, A.; Kangas, H. A Comparative Study of

591

Fibrillated Fibers from Different Mechanical and Chemical Pulps. BioResources 2014, 9

592

(2), 2115–2127.

593

(31) Josset, S.; Orsolini, P.; Siqueira, G.; Tejado, A.; Tingaut, P.; Zimmermann, T. Energy

594

Consumption of the Nanofibrillation of Bleached Pulp, Wheat Straw and Recycled

595

Newspaper through a Grinding Process. Nord. Pulp. Pap. Res. J. 2014, 29 (1), 167–175,

596

DOI 10.3183/NPPRJ-2014-29-01-p167-175.

597

(32) Desmaisons, J.; Boutonnet, E.; Rueff, M.; Dufresne, A.; Bras, J. A New Quality Index for

598

Benchmarking of Different Cellulose Nanofibrils. Carbohydr. Polym. 2017, 174, 318–329,

599

DOI 10.1016/j.carbpol.2017.06.032.

600

(33) Qin, Y.; Qiu, X.; Zhu, J. Y. Understanding Longitudinal Wood Fiber Ultra-Structure for

601

Producing Cellulose Nanofibrils Using Disk Milling with Diluted Acid Prehydrolysis. Sci.

602

Rep. 2016, 6, 1–12, DOI 10.1038/srep35602.

603

(34) Gu, F.; Wang, W.; Cai, Z.; Xue, F.; Jin, Y.; Zhu, J. Y. Water Retention Value for

604

Characterizing Fibrillation Degree of Cellulosic Fibers at Micro and Nanometer Scales.

605

Cellulose 2018, 25 (5), 1–11, DOI 10.1007/s10570-018-1765-8.

606

(35) Hervy, M.; Santmarti, A.; Lahtinen, P.; Tammelin, T.; Lee, K.-Y. Sample Geometry

607

Dependency on the Measured Tensile Properties of Cellulose Nanopapers. Mater. Des.


32

Page 33 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

608


2017, 121, 421–429, DOI 10.1016/j.matdes.2017.02.081.

609

(36) Munier, P.; Gordeyeva, K.; Bergström, L.; Fall, A. B. Directional Freezing of Nanocellulose

610

Dispersions Aligns the Rod- Like Particles and Produces Low-Density and Robust Particle

611

Networks.

612

10.1021/acs.biomac.6b00304.

613 614

Biomacromolecules

2016,

17

(5),

1875–1881,

DOI

(37) Deville, S. Freeze-Casting of Porous Biomaterials: Structure, Properties and Opportunities. Materials (Basel). 2010, 3 (3), 1913–1927, DOI 10.3390/ma3031913.

615

(38) Lavoine, N.; Bergström, L. Nanocellulose-Based Foams and Aerogels: Processing,

616

Properties, and Applications. J. Mater. Chem. A 2017, 5 (31), 16105–16117, DOI

617

10.1039/C7TA02807E.

618

(39) De France, K. J.; Hoare, T.; Cranston, E. D. Review of Hydrogels and Aerogels Containing

619

Nanocellulose.

Chem.

Mater.

620

10.1021/acs.chemmater.7b00531.

2017,

29

(11),

4609–4631,

DOI

621

(40) Varanasi, S.; He, R.; Batchelor, W. Estimation of Cellulose Nanofibre Aspect Ratio from

622

Measurements of Fibre Suspension Gel Point. Cellulose 2013, 20 (4), 1885–1896, DOI

623

10.1007/s10570-013-9972-9.

624

(41) Martinez, D. M.; Buckley, K.; Jivan, S.; Lindstrom, A.; Thiruvengadaswamy, R.; Olson, J.

625

A.; Ruth, T. J.; Kerekes, R. J. Characterizing the Mobility of Papermaking Fibres during

626

Sedimentation. In Proceedings of the Transactions of 12th fundamental Research

627

Symposium, Oxford; 2001; pp 225–254.

628

(42) Gibson, L. J.; Ashby, M. F. Cellular Solids: Structure and Properties; Cambridge


33


629

Page 34 of 37

University Press, 1999; DOI 10.1017/CBO9781139878326.

630

(43) Rezakhaniha, R.; Agianniotis, A.; Schrauwen, J. T. C.; Griffa, A.; Sage, D.; Bouten, C. V.

631

C.; Van De Vosse, F. N.; Unser, M.; Stergiopulos, N. Experimental Investigation of

632

Collagen Waviness and Orientation in the Arterial Adventitia Using Confocal Laser

633

Scanning Microscopy. Biomech. Model. Mechanobiol. 2012, 11 (3–4), 461–473, DOI

634

10.1007/s10237-011-0325-z.

635

(44) Püspöki, Z.; Storath, M.; Sage, D.; Unser, M. Transforms and Operators for Directional

636

Bioimage Analysis: A Survey. In Focus on Bio-Image Informatics; Springer, 2016; pp 69–

637

93, DOI 978-3-319-28549-8_3.

638

(45) Sehaqui, H.; Ezekiel Mushi, N.; Morimune, S.; Salajkova, M.; Nishino, T.; Berglund, L. A.

639

Cellulose Nanofiber Orientation in Nanopaper and Nanocomposites by Cold Drawing. ACS

640

Appl. Mater. Interfaces 2012, 4 (2), 1043–1049, DOI 10.1021/am2016766.

641

(46) Hermans, J. J.; Hermans, P. H.; Vermaas, D.; Weidinger, A. Quantitative Evaluation of

642

Orientation in Cellulose Fibres from the X-Ray Fibre Diagram. Recl. des Trav. Chim. des

643

Pays-Bas 1946, 65 (6), 427–447, DOI 10.1002/recl.19460650605.

644

(47) Apostolopoulou-Kalkavoura, V.; Gordeyeva, K.; Lavoine, N.; Bergström, L. Thermal

645

Conductivity of Hygroscopic Foams Based on Cellulose Nanofibrils and a Nonionic

646

Polyoxamer. Cellulose 2018, 25 (2), 1117–1126, DOI 10.1007/s10570-017-1633-y.

647

(48) Wang, Q. Q.; Zhu, J. Y.; Gleisner, R.; Kuster, T. A.; Baxa, U.; McNeil, S. E. Morphological

648

Development of Cellulose Fibrils of a Bleached Eucalyptus Pulp by Mechanical

649

Fibrillation. Cellulose 2012, 19 (5), 1631–1643, DOI 10.1007/s10570-012-9745-x.


34

Page 35 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


650

(49) Jonoobi, M.; Mathew, A. P.; Oksman, K. Producing Low-Cost Cellulose Nanofiber from

651

Sludge as New Source of Raw Materials. Ind. Crops Prod. 2012, 40, 232–238, DOI

652

10.1016/j.indcrop.2012.03.018.

653

(50) Solala, I.; Volperts, A.; Andersone, A.; Dizhbite, T.; Mironova-Ulmane, N.; Vehniäinen,

654

A.; Pere, J.; Vuorinen, T. Mechanoradical Formation and Its Effects on Birch Kraft Pulp

655

during the Preparation of Nanofibrillated Cellulose with Masuko Refining. Holzforschung

656

2012, 66 (4), 477–483, DOI 10.1515/HF.2011.183.

657

(51) Iwamoto, S.; Abe, K.; Yano, H. The Effect of Hemicelluloses on Wood Pulp

658

Nanofibrillation and Nanofiber Network Characteristics. Biomacromolecules 2008, 9 (3),

659

1022–1026, DOI 10.1021/bm701157n.

660 661

(52) Chinga-Carrasco, G. Optical Methods for the Quantification of the Fibrillation Degree of Bleached MFC Materials. Micron 2013, 48, 42–48, DOI 10.1016/j.micron.2013.02.005.

662

(53) Henriksson, M.; Berglund, L. A.; Isaksson, P.; Lindström, T.; Nishino, T. Cellulose

663

Nanopaper Structures of High Toughness. Biomacromolecules 2008, 9 (6), 1579–1585,

664

DOI 10.1021/bm800038n.

665 666

(54) Deville, S.; Saiz, E.; Nalla, R. K.; Tomsia, A. P. Freezing as a Path to Build Complex Composites. Science (80-. ). 2006, 311 (5760), 515–518, DOI 10.1126/science.1120937.

667

(55) Wegst, U. G. K.; Schecter, M.; Donius, A. E.; Hunger, P. M. Biomaterials by Freeze

668

Casting. Philos. Trans. R. Soc. London A Math. Phys. Eng. Sci. 2010, 368 (1917), 2099–

669

2121, DOI 10.1098/rsta.2010.0014.

670

(56) Gindl, W.; Keckes, J. Drawing of Self-Reinforced Cellulose Films. J. Appl. Polym. Sci.


35


671

Page 36 of 37

2007, 103 (4), 2703–2708, DOI 10.1002/app.25434.

672

(57) Glicksman, L. R. Heat Transfer in Foams. In Low density cellular plastics; Hilyard, N. C.,

673

Cunningham, A., Eds.; Springer, Dordrecht, 1994; pp 104–152, DOI 10.1007/978-94-011-

674

1256-7_5.

675

(58) Collishaw, P. G.; Evans, J. R. G. An Assessment of Expressions for the Apparent Thermal-

676

Conductivity of Cellular Materials. J. Mater. Sci. 1994, 29 (9), 2261–2273, DOI

677

10.1007/BF00363413.

678 679 680 681

(59) Berge, A.; Johansson, P. Ä. R. Literature Review of High Performance Thermal Insulation; Chalmers University of Technology, 2012. (60) Zeng, S. Q.; Hunt, A.; Greif, R. Transport Properties of Gas in Silica Aerogel. J. Non. Cryst. Solids 1995, 186, 264–270, DOI 10.1016/0022-3093(95)00052-6.

682 683 684


36

Page 37 of 37 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


685

SYNOPSIS

686

Correlating the porous structure, the compressive behaviour and the cellulose nanofibril (CNF)

687

alignment in the cell walls of anisotropic CNF foams enables assessment of the nanofibrillation

688

quality.

689

TOC/Abstract graphic

690

For Table of Contents Use Only

691 692


37

Analysis of the porous architecture and properties ... - ACS Publications

Recommend Documents