Nanofibrillated Cellulose Templated Membranes with High

Nov 18, 2016 - ... index (PdI) were determined according to the ISO standard 22412.(60) The ...... (Cambridge, U. K.) 2013, 49 (83) 9564– 9566 DOI: ...
0 downloads 0 Views 2MB Size
Subscriber access provided by University of Otago Library

Article

Nanofibrillated cellulose templated membranes with high permeance Paola Orsolini, Tommaso Marchesi D'Alvise, Cristiana Boi, Thomas Geiger, Walter Remo Caseri, and Tanja Zimmermann ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b12107 • Publication Date (Web): 18 Nov 2016 Downloaded from http://pubs.acs.org on November 20, 2016

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 free 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 accessible to all readers and 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.

ACS Applied Materials & Interfaces 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 42

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 Applied Materials & Interfaces

1

Nanofibrillated cellulose templated membranes with

2

high permeance

3

Paola Orsolini,a,c* Tommaso Marchesi D’Alvise,a,b Cristiana Boi,b Thomas Geiger,a Walter R.

4

Caseri,c Tanja Zimmermanna*

5

a

6

Materials Laboratory, Überlandstrasse 129, CH-8600 Dübendorf, Switzerland.

7

b

8

c

Empa – Swiss Federal Laboratories for Material Science and Technology, Applied Wood

DICMA, Alma Mater Studiorum-Università di Bologna, via Terracini 28, 40131 Bologna, Italy

ETH Zürich, Multifunctional Materials, Vladimir-Prelog-Weg 5, 8093 Zürich, Switzerland

9 10

KEYWORDS

11

calcium compounds, nanofibrillated cellulose membranes, template, permeance, filtration.

12 13

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces

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 42

1

ABSTRACT

2

One of the most challenging aspects of using nanofibrillated cellulose (NFC) for membranes

3

production is their limited permeance. When NFC membranes are produced from aqueous

4

suspensions, depending on their grammage, the permeances are in the range of a few decades of

5

L/(hm2MPa) not matching satisfactory filtration times. We present a fast and sustainable solution

6

to increase the permeances of such membranes through a combination of solvent exchange of the

7

NFC suspension with ethanol and the use of a removable template, a mixture of calcium

8

compounds (CC). The effect of the CC/NFC ratio was screened for various concentrations. The

9

permeance of water could be increased by as much as 2-3 times as compared to non-templated

10

membranes. Further, the membranes showed the ability for penetration of water-soluble

11

macromolecules, contaminant rejection of suspended solid particles and thus fluids (such as

12

orange juice) could be concentrated, with a view to applications in food industry.

13 14

1 INTRODUCTION

15

Nanofibrillated cellulose (NFC) is a bio-based material derived from pristine and waste

16

cellulose fibers through water-based disintegration processes such as grinding or high shear

17

homogenization

18

been developed. While initially the mechanical and barrier properties of NFC in polymeric

19

composite materials

20

use of NFC for environmental remediation

21

Carpenter et al.

1–4

22

. Since NFC was first described by Turbak et al. 5, several applications have

3,6–11

have been exploited, more recently, many works have dealt with the 12–14

and for preparation of membranes

15–21

. Indeed,

emphasized the potential of nanocellulose materials such as NFC, for waste-

ACS Paragon Plus Environment

2

Page 3 of 42

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 Applied Materials & Interfaces

1

water treatment and purification purposes. Moreover, recent studies on its life cycle assessment

2

showed the interest of using NFC compared to other nanomaterials, such as carbon nanotubes23.

3

Originally, NFC active layers were applied on various polymer substrates

24–28

to obtain

4

composite membranes with high flux and permeance. Nowadays new approaches for membranes

5

preparation from NFC or NFC derivatives are envisioned to substitute synthetic polymers. In

6

fact, compared to membranes made of regenerated cellulose and its derivatives, NFC-based

7

membranes can offer at the same time an active layer for filtration and a mechanically stable

8

support layer; in addition, the strength and stiffness properties are provided by hydrogen bonds

9

acting among the entangled cellulose nanofibers. The densely packed structure of NFC with 29

10

pores mostly in the mesoporous (2-50 nm) and macroporous (> 50 nm) range

makes such

11

materials suitable for ultrafiltration operations. At present nanocellulose derived membranes

12

have been prepared for the rejection of ions 16, adsorption of heavy-metal ions 17,30, adsorption of

13

nitrates 31 and organic compounds such as humic acid 32, purification of solvents 15, hemodialysis

14

19

, extraction of DNA and oligomers 20 and virus removal 21.

15

The amphiphilic behavior of cellulose and the presence of hydroxyl groups on the nanofibers

16

surface allow for chemical functionalization. Hydrophobization strategies by esterification and

17

silylation were extensively studied for nanocellulose to promote water repellency and enhancing

18

the affinity for oily phases

19

membranes suitable for oil from oil-in-water emulsion separations.

12,33–38

. Such protocols might be applied for preparing NFC-based

20

The essential factor for membrane applications based on the size-exclusion mechanism is their

21

pore size that in the case of NFC depends on the NFC network arrangement. Therefore, it has to

22

be taken into account that during NFC membrane preparation from aqueous suspensions, the

23

nanofibers frequently collapse into a densely packed network. In these cases low permeances in

ACS Paragon Plus Environment

3

ACS Applied Materials & Interfaces

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

1

the range of few decades of L/(hm2MPa) were measured

2

the water in NFC dispersions could be exchanged by organic solvents, thus allowing a faster

3

drying process suppressing a tight contact of the fibrils

4

product possesses lower mechanical stability than the one prepared from water, as the number of

5

hydrogen bonds decreases due to larger distances of the nanofibrils. The possibility to control the

6

variation of membranes pore sizes is clearly of great importance for filtration technologies. Often

7

hard- and soft templating methods are reported to introduce pores with uniform shape and size in

8

continuous solid phases and in membranes

9

templated polymeric membranes of polyether sulfone (PES) and cellulose acetate (CA) based on

10

the use of calcium carbonate and other nanoparticles 50,51. Thus membranes for ultrafiltration and

11

dialysis were produced, respectively, and the implementation of the template removal into an

12

industrial roll-to-roll process was shown 52.

43–49

15,16

Page 4 of 42

39–42

. To prevent nanofiber collapsing,

. On the other hand, the final dry

. To this aim, Kellenberger et al. developed

13

In this work, we present a green method to increase the permeance of NFC membranes by a

14

combination of solvent-exchange and templating approach. Calcium compounds (a mixture of

15

species such as carbonates, hydrogen carbonates and hydroxides) (CC) particles were introduced

16

into, and subsequently removed from, NFC membranes to confer additional pores to the

17

membrane substrate and thus to modify the membrane properties. The possibility to achieve an

18

in-situ CC formation on NFC in ethanol was also briefly explored.

19 20

2

21

2.1 Materials

22

Never-dried cellulose pulp (Elemental Chlorine Free grade) composed of spruce and pine fibers

23

was provided by Zellstoff Stendal (Germany). Ethanol reagent grade and ammonium hydrogen

EXPERIMENTAL

ACS Paragon Plus Environment

4

Page 5 of 42

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 Applied Materials & Interfaces

1

carbonate were purchased from VWR Chemical (Belgium). Calcium chloride dihydrate p.A.,

2

calcium carbonate p.A., ethylenediamine tetraacetate (EDTA) aq. solution of 0.1 mmol/L and

3

acetic anhydride were purchased from Merck KGaA (Germany). The calcium-sensitive indicator

4

calconcarboxylic acid, Merck KGaA (Germany) was mixed with sodium chloride, VWR

5

Chemical (Belgium), to produce a mixture (0.05 g of indicator into 5 g of NaCl) to be used

6

during back-titration with EDTA. Potassium hydroxide was purchased from Fluka (Switzerland).

7

Polyethylene glycol (PEG) of different molecular masses was purchased from various providers:

8

PEG 35,000 g/mol from Fluka (Germany), 100,000 g/mol from Alfa Aesar, 200,000 g/mol and

9

5,000,000 g/mol from Polyscience Inc. (USA) and 900,000 g/mol from Janssen Chimica

10

(Belgium). Carbon black powder was obtained from Columbian Chemicals Canada Ltd.

11

(Canada). Alfa Laval (Denmark) kindly provided industrial ultrafiltration membranes made of

12

regenerated cellulose acetate (RC).

13 14 15

2.1.1

NFC Suspension Preparation

16

NFC was prepared in water from swollen pulp fibers by grinding for four passes and

17

homogenizing for six passes as reported elsewhere 29. A final dry matter content of 1.6 wt% after

18

a total number of 10 passes through the process was thus obtained. A certain amount of NFC

19

suspension in water was taken apart and used for the preparation of dense NFC membranes. The

20

remaining part of the water suspension was solvent-exchanged with ethanol for the preparation

21

of templated membranes with CC particles. The exchange was achieved by centrifuging for five

22

times at 5,000 rpm for 20 min each cycle and by adding 45 mL of fresh ethanol, alternating

23

dispersion steps performed for 5 min at 10,000 rpm by a Digital Ultraturrax T25 (Ika, Germany).

ACS Paragon Plus Environment

5

ACS Applied Materials & Interfaces

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 6 of 42

1 2

2.1.2

Synthesis of CC Particles

3

Stable CC particles were synthesized, in ethanol, based on the protocol developed by Chen et al.

4

53

5

three days at room temperature. Open glass bottles containing a solution of 100 mL of ethanol

6

and 200 mg of solid calcium chloride dihydrate, obtained by stirring for 30 min at 250 rpm at

7

room temperature, were placed in the desiccator near an open container of solid ammonium

8

hydrogen carbonate. After three days, the formation of blue suspended solids inside the bottles

9

was attained. As the final product most likely consists of a mixture of different calcium

10

compounds in equilibrium, such as, calcium carbonate, calcium hydrogen carbonate and calcium

11

hydroxide, we define here for simplicity the products as calcium compounds “CC”.

12 13 14

2.1.3

15

All membranes were prepared by a paper-making approach. Dense NFC membranes with

16

grammage (GR) of 30 g/m2 and 100 g/m2 were first prepared from a water suspension as

17

presented by Mautner et al. 15. In parallel NFC membranes with GR 30 g/m2 and 100 g/m2 were

18

prepared after solvent exchange to ethanol. Finally, membranes of NFC in ethanol were prepared

19

based on the exo-template approach, by dispersing the CC particles in the starting ethanol

20

suspension of nanofibers. Unblended and blended systems with CC template were first dispersed

21

in ethanol by a Digital Ultraturrax T25 for 5 min at 10,000 rpm (Ika, Germany), outgassed under

22

vacuum for 20 min to remove air bubbles trapped during mixing and finally vacuum filtered to

23

obtain a wet cake. This was later dried by hot-pressing at 105 °C for 20 min, under an applied

24

pressure of 40 kPa. Assuming a 100 % reaction yield for the formation of CC and under the

25

assumption that no losses during filtration occur, three different theoretical ratios of grams of

. The synthesis was performed in a desiccator under reduced pressure of about 0.1 mbar for

Exo-templated and Endo-templated CC Membranes Preparation and Template Removal

ACS Paragon Plus Environment

6

Page 7 of 42

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 Applied Materials & Interfaces

1

template over dry mass of NFC material were considered: 1/50 (designated as low ratio L), 1/10

2

(medium ratio M) and 1/5 (high ratio H) (detailed preparation in SI, section S2.1). Membrane

3

thicknesses were measured by means of a caliper (Futuro, Switzerland) by considering at least

4

three membranes per type and 35-40 measurements per membrane. CC particles were removed

5

by washing with HCl solution at 0.1 M and 1 M for 30 min and 1 M for 60 min, respectively.

6

These different conditions were investigated to optimize the sacrificial template removal. After

7

template removal, the membranes were rinsed with water until neutral pH was reached. In other

8

experiments, the reaction for the synthesis of calcium carbonate was executed in presence of ~

9

0.3 g dry mass NFC suspended in 100 mL ethanol to induce in-situ growth of calcium carbonate,

10

also defined here as an endo-templated method.

11 12

2.2 Methods

13

2.2.1 Scanning and Transmission Electron Microscopy

14

Scanning Electron Microscopy (SEM) was executed using a FEI Nova NanoSEM 230

15

Instrument (Fei, USA), at an accelerating voltage of 5 kV. Drops of 0.1 wt% of NFC in ethanol

16

suspension were placed on a mica substrate and sputtered with a 7 nm platinum layer. SEM was

17

utilized to evaluate pulp fibers and nanofibers for which length (L) and diameter (D) were

18

determined. Transmission electron microscopy (TEM) was performed with a field emission

19

transmission electron microscope Jeol JEM-2200FS (Jeol USA Inc., USA), which was utilized at

20

200 kV to observe CC particles and to evaluate their crystallinity by selected area diffraction

21

patterns (SAED-TEM). The particles were dispersed in ethanol at a concentration of ~0.01 wt %

22

and a drop was deposited on a copper TEM grid and evaporated at room temperature. The image

23

evaluation was carried out by means of ImageJ 54.

ACS Paragon Plus Environment

7

ACS Applied Materials & Interfaces

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

Page 8 of 42

1

2.2.2 Thermogravimetric Analysis (TGA)

2

For TGA measurements a TGA 7 (Perkin Elmer, USA) was used with a scanning speed of 20

3

°C/min under helium atmosphere from 25 °C to 900 °C.

4

2.2.3 Dynamic Light Scattering (DLS)

5

DLS measurements were executed by a Zetasizer Nano ZS (Malvern instruments, USA).

6

Measurements were carried out on 2 mL volume of suspension after mixing at room temperature

7

in duplicates for each reaction to exclude sampling errors. Dynamic light scattering (DLS),

8

which provides the hydrodynamic diameter (DH) of particles and polydispersity index (PdI)

9

was employed for the characterization of the CC particles dispersed in ethanol. DH corresponds

10

to the diameter of an equivalent solid sphere surrounded by a few layers of wetting liquid

11

subjected to the same friction of the real particle itself. The average size and the polydispersity

12

index (PdI) were determined according to the ISO standard 22412

13

number between 0 and 1, for distribution with PdI above 1 the DLS technique is not applicable

14

due to the wide broadness of the size distribution.

15

2.2.4 Determination of the Total Calcium Content in the NFC Membranes

16

The determination of the inorganic calcium compound content was carried out by means of back-

17

titration

18

template removal were charred in a muffle oven at 575°C ± 5 °C for 3 h. The ashes were

19

dispersed first in 3 mL of deionized water and 0.5 mL of aq. HNO3 (65 %) solution and

20

intensively mixed. Subsequently, the solution was diluted to a total volume of 100 mL in

21

deionized water (for samples with low and medium CC concentrations) and to 250 mL in

22

deionized water (for samples with high CC concentration). Aliquots of 20 mL were separated

23

and their pH adjusted to 12 adding dropwise a 17.8 M aq. KOH solution. Finally, the solution

55,56

60

59

,

. The PdI is defined as a

. First, the NFC membranes containing the CC template and the samples after the

ACS Paragon Plus Environment

8

Page 9 of 42

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 Applied Materials & Interfaces

1

was titrated by means of a 0.01 M aq. EDTA solution. A mixture of calcium-specific indicator

2

calconcarboxylic acid and NaCl (masses 0.02-0.03 g) was used as aid to detect the color change

3

of the solution from violet to blue. The results provided the concentration of Ca2+ ions of various

4

calcium compounds, such as calcium carbonates, calcium hydrogen carbonates and calcium

5

hydroxides, and were directly related to the equivalent amount of CaCO3 present in the

6

membrane preparation. The technique was first assessed by establishing a curve by titrating

7

known quantities of CaCO3 (SI, section S3.1).

8

2.2.5

9

ATR-IR spectra were acquired with a Tensor 27 (Bruker, Germany).

Fourier Transformed Infrared – Attenuated Total Reflectance Spectroscopy (ATR-IR)

10

2.2.6 X-ray Diffraction (XRD)

11

X-ray analysis was performed using a Bruker D8 Advance (Bruker) with Cu Kα radiation (λ =

12

1.54 Å, a voltage of 40 KV and a current of 40 mA, parallel beam), scanning from 5 ° to 50 °.

13

The crystallinity index (Ic) was calculated according to Eq. 1 established by Segal et al. 57:

14

𝑰𝒄 = (𝟏 −

𝑰𝟏 𝑰𝟐

)

(1)

15

where I1 and I2 are, respectively, the minimum peak intensity at 2θ = 18° and the peak intensity

16

given by the crystalline region of the cellulose 2θ = 22.5°.

17

2.2.7 Pore Size Distribution in Dry NFC Templated Membranes

18

Mercury intrusion was carried out on dry templated membranes. After template removal the

19

membranes were flushed with ethanol for 3 times, soaked one more time in ethanol for 30 min,

20

and dried at 70 °C for 2 hours under a load. Afterwards the membranes were cut into small

21

pieces (size range mm x mm) and dried again at 70 °C for 2 hours to remove adsorbate prior to

22

the mercury intrusion. Pascal 140 and Pascal 440 mercury Porosimeters (Thermo Fisher,

ACS Paragon Plus Environment

9

ACS Applied Materials & Interfaces

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

Page 10 of 42

1

Germany) were employed to evaluate the pore size distribution of the membranes developed in

2

this work. A two-step method comprised one measurement at low (up to 120 kPa) and one at

3

high intrusion pressure (up to 350 MPa), both at room temperature. The mercury density was

4

automatically corrected at the actual temperature. The values of mercury contact angle and the

5

surface tension were respectively fixed at 140° and 0.48 N/m. Samples were run in duplicates

6

and the final curves averaged.

7 8

2.2.8 Water Permeance Tests and Permeability

9 10

Permeability tests were carried out in a high-pressure filtration device HP4750 (Sterlitech,

11

USA). A dead-end configuration set-up was employed, connected to a pressurization system of

12

inert gas that allowed screening pressures between 0.5 and 3.5 bar, by 0.5 bar step increase every

13

15 min. The measuring cell resistance to the flow is negligible. The membranes were cut into

14

circles with a 50 mm diameter, swollen in deionized water (pH 6.4, Ω ~ 8 μS/cm) for 48 h before

15

measurement to ensure an almost complete swelling of the cellulosic material and subjected to

16

filtration of 150 mL of deionized water. The mass of permeate was recorded over time with a

17

Mettler MS204S (Mettler, Switzerland) balance, connected to an automatic data acquisition

18

software. The permeance was calculated dividing the water flux (Q in L/h) by the active area of

19

the membrane (A) (in m2) by the pressure ΔP (Pa) applied during filtration. The membrane

20

hydraulic permeability (Lp in m) was calculated from Darcy’s law 58, as reported in Eq. (2):

21 22

𝑸 𝑨

= 𝑳𝒑

𝜟𝑷 𝝁

(2)

in which μ is the dynamic viscosity of water (Pa s).

ACS Paragon Plus Environment

10

Page 11 of 42

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 Applied Materials & Interfaces

1 2 3

2.2.9

4

In a first experiment, polyethylene glycol (PEG) solutions of different PEG molecular masses in

5

demineralized water at a constant concentration of 1 g/L were utilized to assess the filtration

6

properties of the membranes. Filtration experiments were performed using an Amicon® stirred

7

cell, model 8200, of 200 mL volume (Millipore, USA), filled with 30 mL of PEG solutions,

8

filtering 20 mL of permeate, applying a pressure of 2 bar and operating at 200 rpm. The same

9

membrane was used for a maximum of 3 experiments, cleaning the membrane and the cell after

10

each filtration. The system was cleaned by washing with 40 mL of water at 100 rpm for 2

11

minutes and then by filtering of 40 mL of demineralized water at 3 bar and 100 rpm. A feed

12

sample, a retentate sample and a permeate sample were collected for each measurement. The

13

membrane rejection (R %), is defined as in Eq. 3:

14

Filtration of Water-soluble Macromolecules, Particle Separation and Concentration of Fluids

𝑹 % = (𝟏 −

𝑪𝒑 𝑪𝒇

) × 𝟏𝟎𝟎

(3)

15

where Cp is the concentration of the solute in the permeate and Cf is the concentration of the

16

solute in the feed. The permeate and the feed samples were analyzed using a HPLC system with

17

refractive index detector, model RID-10A (Shimadzu, Japan) at 37 °C, by-passing the column

18

and using a flow rate of 0.1 mL/min of demineralized filtered water as mobile phase, and

19

injecting 100 µL of sample solution. The areas which correspond to sample concentrations were

20

calculated by integrating the peaks with the HPLC software Lab Solution. In order to estimate

21

the ability of templated membranes to concentrate fluids containing dissolved molecules and

22

suspended solids, commercially available orange juice (Andros, Switzerland) with a starting

23

solid content of 12 wt% was filtered through NFC templated membranes with grammage 100

24

g/m2. A high-pressure filtration device HP4750 (Sterlitech, USA) was utilized by applying a

ACS Paragon Plus Environment

11

ACS Applied Materials & Interfaces

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

Page 12 of 42

1

pressure of 3.5 bar on 75 mL of juice. For each sample, one aliquot of permeate was collected

2

after four hours and its refractive index measured at 25 °C by an Abbe Refractometer (Zeiss,

3

Germany). The clarity of permeates was determined in transmittance (T) by means of a

4

spectrophotometer Cary 50 Scan (Varian, USA). For tests with carbon black particles, the carbon

5

black particles were dispersed in water by sonication for 2 min. A feed concentration of 0.05 g/L

6

was utilized as probing material for the determination of the solid removal efficiency of

7

templated membranes. The particle size distribution of carbon black was evaluated in water by

8

means of laser scattering with a LS 13 320 Particle Size Analyser (Beckman Coulter, USA) at an

9

obscuration value between 6-12 % (results are presented in the SI, section SI 4.1). The feed (75

10

mL) was filtered through NFC templated membranes of grammage 30 g/m2 and through an

11

ultrafiltration regenerated cellulose acetate membrane (RC) used as benchmark, in a high-

12

pressure filtration device HP4750 (Sterlitech, USA), operated at 3.5 bar. Carbon black

13

concentration in the permeate was evaluated by means of a spectrophotometer Cary 50 Scan

14

(Varian, USA) measuring the transmittance (T).

15

3

16 17 18

3.1 Characterization of the Membranes Components and Membrane Preparation with an Exo-templating Approach.

19

The starting cellulose pulp, composed of fibers with lengths in the order of a few millimeters and

20

diameters in the order of 30 μm (Figure 1a) was disintegrated into fibrils with aspect ratios (L/D)

21

above 50. The shape and the size of CC particles are shown (Figure 1c). We assume that a

22

mixture of calcium carbonate, calcium hydrogencarbonate and calcium hydroxide is present due

23

to equilibria with CO2 and water. The formation of an amorphous material was attained as

RESULTS AND DISCUSSION

ACS Paragon Plus Environment

12

Page 13 of 42

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 Applied Materials & Interfaces

1

presented in Figure 1d. Figure 1f shows the particle size distribution of CC particles for several

2

replicates. The average size and the polydispersity index (PdI) amounted to 135 ± 21 nm (95 %

3

confidence level), with a PdI of 0.022 ± 0.003 (95 % confidence level). The distributions were

4

well reproducible. The particles were also characterized after drying in an oven, since thermal

5

treatment can change the composition of CC phases. As such, in our case by oven-drying the CC

6

from ethanol, the formation of crystalline phases slightly different from the ones achieved in our

7

work could be induced. A detailed description of dry powders is provided in the supplementary

8

information (SI, section 1.1, Figure S1 and S2), including FTIR spectroscopy analysis and

9

thermal analysis (TGA). Both methods can be applied on dry powders but not on aqueous

10

suspensions of CC particles. Since the membrane preparation involved mixing of NFC and CC

11

followed by a thermal treatment by hot-pressing, we should expect a mixture of calcium

12

compounds of various composition in the templated membranes, which might have a relevance

13

on the results presented in section 3.3 related to the CC dissolution.

ACS Paragon Plus Environment

13

ACS Applied Materials & Interfaces

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

Page 14 of 42

1 2

Figure 1 a) SEM micrograph of starting softwood pulp fibers (scale bar 100 μm); b) SEM

3

micrograph of NFC obtained from pulp fibers by a combined grinding- and high pressure

4

homogenization process after ten processing cycles (scale bar 100 μm); c) TEM image of

5

calcium-based particles (CC) synthesized in ethanol (scale bar 0.5 μm); d) SAED-TEM of CC

6

particles showing an amorphous phase formation (insert image, scale bar 100 nm); e) diameter

7

size distribution of CC particles derived by image analysis of TEM micrographs, and f) particle

8

size distribution of CC particles determined by dynamic light scattering measurements showing

9

an average size of 135 ± 21 nm (95 % confidence level) and a polydispersity index (PdI) of

10

0.022 ± 0.003 (95 % confidence level).

11

ACS Paragon Plus Environment

14

Page 15 of 42

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 Applied Materials & Interfaces

1

The produced CC particles (suspension in ethanol) and an ethanolic NFC suspension were mixed

2

together fixing the cellulose grammage of the membranes at 30 and 100 g/m2 following the

3

principle of an exo-templating approach 48. In order to investigate the effect of different template

4

loadings on the final properties of the NFC membranes, the CC content was varied to provide

5

low (L), medium (M) and high (H) theoretical ratios of CC/NFC (GR30 L, GR30 M, GR30 H,

6

GR100 L, GR100 M and GR100 H) (detailed list of samples and relative quantities are reported

7

in SI, section 2.1). The effective CC/NFC ratios might differ somewhat from the theoretical

8

values because of losses of CC during the membrane making process. A precise determination of

9

the calcium content was later determined in paragraph 3.3.

10 11 12 13

3.2 In-situ Formation of CC on NFC: One-step Production of Endo-templated Membranes

14

CC particles were also synthesized in presence of NFC in ethanol suspension to investigate the

15

feasibility of a single-step approach, compared to the combination of two separated ethanol-

16

suspensions described above (exo-templating). In the single-step process the template grows

17

inside its own matrix, fulfilling an endo-templating approach 48. A similar idea was presented by

18

Mohammadkazemi et al. to include calcium carbonate directly during bacterial cellulose

19

synthesis to produce novel biocomposites

20

section 3.1, a difference in the size of the particles is immediately evident (particles of sizes

21

always between 50 nm and 180 nm, Figure 2) and the polydispersity of the CC particles

22

appeared higher (cf., Figure 1d). The particles seemed to adhere at the surface of the nanofibers

23

and in some cases filled the interspace among them, showing the existence of attractive forces

24

between cellulose and CC particles also in presence of ethanol. The determination of

61

. Compared to the exo-templating methods in

ACS Paragon Plus Environment

15

ACS Applied Materials & Interfaces

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

Page 16 of 42

1

hydrodynamic radii (DH) by DLS was not possible, as the presence of the nanofibers interfered

2

with the particles signals. Hence, we preferred to prepare template membranes by the exo-

3

templating approach.

4

5 6

Figure 2 In-situ grown CC particles on NFC (scale bar 500 nm).

7 8

3.3 Removal of Calcium-based Compounds

9 10

The CC particles were later removed by acidic washing with aqueous HCl solution under

11

formation of Ca2+ ions and carbon dioxide

12

investigated to define the most suitable HCl concentration and the contact time for dissolution to

13

minimize degradation of the cellulose. Membranes with GR 100 g/m2 and at high ratio of

14

CC/NFC (samples named GR100 H), were evaluated and all results in this section refer to those.

15

In a first series of experiments, sequential washings were carried out on the same membrane as

16

follows: 0.1 M aq. HCl solution for 30 min, 1 M aq. HCl solution for 30 min and 1 M aq. HCl

52

. The optimal condition for the CC removal was

ACS Paragon Plus Environment

16

Page 17 of 42

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 Applied Materials & Interfaces

1

solution for 60 min to evaluate whether by using harsher conditions, a larger amount of CC

2

template would be removed. The water permeance was measured at each step. Figure 3 shows

3

the results of water permeance measurements for different washings, compared to a reference

4

membrane of GR 100 g/m2 prepared from ethanol.

5 6

Figure 3 Water permeance of a NFC membrane of GR 100 g/m2 (dark blue curve) compared

7

with templated membranes prepared with high ratio CC/NFC (GR100 H) subjected to a

8

sequential washing with different HCl concentrations and times: 0.1 M for 30 min in light blue, 1

9

M for 30 min in purple and 1 M for 60 min in brown, respectively.

10

After about 100 – 150 min the systems reached steady state and the average permeance values

11

were calculated: for the reference membrane prepared from ethanol without template, the steady

12

state value of permeance was 51 L/(hm2MPa). In the case of templated membranes, the water

13

permeance increased to 71 L/(hm2MPa) after a relatively mild washing (0.1 M aq. HCl solution

14

for 30 min) and to 78 L/(hm2MPa) after a more severe washing (1 M aq. HCl solution for 30

15

min). No significant differences in permeation were observed for longer washing times (78

16

L/(hm2MPa) at 1 M aq. HCl solution for 60 min). The permeance results suggested that cleaning

ACS Paragon Plus Environment

17

ACS Applied Materials & Interfaces

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

Page 18 of 42

1

the membranes at 1 M aq. HCl solution for 30 min was sufficient to remove the CC from the

2

NFC matrix. Normally, amorphous calcium carbonate phases undergo faster dissolution

3

compared to their crystalline polymorphs; furthermore, the particles own high specific surface

4

energy, thus they exist in an energetically metastable state, rendering their dissolution favored 62.

5

Regarding the stability towards pH, it is known that different pH values can change the swelling

6

ability of cellulose materials especially in strong alkaline conditions. However, in our case low

7

pH was used, thus this should exert minor influence on the 3D arrangements of pores. After

8

template removal the templated membranes were washed until neutral pH was reached in order

9

to evaluate the structure in the same conditions.

10

SEM microscopy was used to visualize the surface topography of membranes (GR100 H) before

11

and after CC removal (Figure 4). Before washing with aq. HCl solution, CC particles here

12

delineated as white round beads, were randomly and homogeneously dispersed and did not form

13

agglomerates neither at the surface of the membrane (Figure 4a1) nor in the bulk (cross section,

14

Figure 4a2). The blending with NFC did not considerably change their sizes, since particles in the

15

range of approximately 100-150 nm diameters were always observed. After washing, only a few

16

residual particles were present (Figure 4b1-2) and the structure of the membranes appeared more

17

open. Round voids (negative shape of the particles) were not discerned. We could not claim that

18

newly introduced voids were round, since the structure visualized in the cross section is always

19

influenced by the samples preparation.

20

ACS Paragon Plus Environment

18

Page 19 of 42

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 Applied Materials & Interfaces

1 2

Figure 4 SEM images of membranes of GR 100 g/m2 prepared from ethanol: a1-2) top view and

3

cross view, respectively, of membrane GR100 H before the template removal; b1-2) top view and

4

cross view, respectively, of the same membrane after 30 min washing with 1 M aq. HCl solution

5

(red arrows pointing at residual CC particles).

6

ACS Paragon Plus Environment

19

ACS Applied Materials & Interfaces

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

Page 20 of 42

1 2

Figure 5 Evaluation of the removal efficiency of calcium phases of NFC membranes of GR 100

3

g/m2 templated with high ratio of CC/NFC (GR100 H) compared to the behavior of a NFC

4

membrane of GR 100 g/m2 prepared from ethanol (blue line), a NFC membrane without

5

template prepared from ethanol and subjected to an acidic washing (black line), compared to a

6

templated membrane GR100 H CC/NFC unwashed (red line) and washed with 1 M aq. HCl

7

solution after 30 min (green line). a) FTIR spectra and b) XRD patterns; c1-2) thermograms

8

showing the mass-loss derivative and the mass loss for the aforementioned samples, respectively.

9

FTIR spectra (Figure 5a) indicate that calcium compounds, detected at wavelengths between

10

1600-1400 cm-1 (red continuous line), disappear after washing with 1 M aq. HCl solution for 30

11

min. X-Ray diffraction (Figure 5b) was utilized to investigate possible differences in the

ACS Paragon Plus Environment

20

Page 21 of 42

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 Applied Materials & Interfaces

1

crystallinity of cellulose and CC particles upon washing. Basically, acidic treatments can lead to

2

the dissolution of cellulose amorphous domains in the nanofibers; however the extent of the

3

dissolution by using HCl depends on the concentrations, dissolution temperature and exposure

4

times

5

Eq. 1, for the characteristic peak intensities at 2θ of 18° and 22.5° (Table 1). No significant

6

differences were observed before and after the acidic washing, as Ic varied between 0.70 and

7

0.74. From the patterns (Figure 5b) it was not possible to observe the presence of CC in the

8

sample with high ratio CC/NFC (GR100 H), due to broad signals.

9

Table 1 Crystallinity Index (Ic) determined with Equation 1, based on two characteristic

10

63

. From the XRD patterns the cellulose crystallinity index (Ic) was calculated based on

cellulose peaks at 2θ of 18 ° and 22.5 °. I1 (2θ = 18°)

I2 (2θ = 22.5°)

Ic

NFC mem

22.42

75.67

0.70

NFC mem washed

16.86

60.23

0.72

High ratio CC/NFC unwashed

10.85

42.21

0.74

High ratio CC/NFC washed

15.61

57.02

0.73

11 12

Thermogravimetric analysis (TGA) confirmed the removal of CC by washing with 1 M aq.

13

HCl solution for 30 min. Figure 5c1-2 displays the mass-loss derivative dependence on the

14

temperature for a sample with Gr 100 g/m2 (GR100 H CC/NFC), compared to the corresponding

15

membrane without CC (same grammage, not treated with acid and pre-washed at the same acidic

16

conditions as the CC-containing membrane). The onset temperature (Ton) of mass loss attributed

17

to sample degradation arose around 340 °C for the membrane without CC and the membrane

18

after washing, while Ton around 307 °C was found in presence of CC. A peak at 651 °C was

19

observed most probably due the conversion of CaCO3 to CaO and condensation of Ca(HCO3)2 64.

ACS Paragon Plus Environment

21

ACS Applied Materials & Interfaces

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

Page 22 of 42

1

In Figure 5c2 a final residue at 900 °C of 3.1 % for the unwashed sample also confirmed the

2

presence of the inorganic template. Such a residue did not emerge in the TGA of the CC/NFC

3

membrane subjected to the acidic washing (residual mass at 900 °C < 1%) which matched the

4

degradation curve of the cellulosic reference membrane. Thus, the thermograms of the samples

5

without CC and the templated membranes after washing were very similar but clearly

6

distinguished from the samples with CC before washing. Overall, we can state that from the

7

aforementioned FTIR spectra and TGA measurements, a washing with 1 M aq. HCl solution for

8

30 min essentially removed the calcium compounds, within the detection limits of the

9

techniques, as already evident from the SEM investigations. Furthermore, no significant

10

degradation of the cellulose occurred as indicated by the constant values of the crystallinity index

11

and the FTIR spectra. Since the optimization was carried out on the membrane containing the

12

maximum loading of CC among all samples investigated, we assume that the washing is also

13

efficient for samples containing less CC mass. Accordingly, this washing condition was kept for

14

all the subsequent investigations with CC-containing membranes (section 3.4).

15 16

3.4 Determination of Calcium Content by EDTA Titration

17 18

Although the FTIR spectra, XRD patterns and TGA traces did not disclose the presence of

19

calcium compounds after template removal on the samples owing higher CC/NFC ratio, the

20

actual calcium content was determined by complexometric titration (Supplementary Information,

21

Figure S3 and S4), which is more sensitive. Complexometric titration by EDTA was already

22

applied on calcium carbonate-containing paper

23

mmolCC/gPaper

56

55

, and showed detectability down to ca. 0.0001

. The content in milligrams of calcium was calculated after measuring the

ACS Paragon Plus Environment

22

Page 23 of 42

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 Applied Materials & Interfaces

1

equivalent content of Ca2+ inside the membranes. The assessment of Ca2+ ions was carried out by

2

means of a back-titration with EDTA after charring of cellulose and dissolution of the ashes into

3

acidic solutions of HNO3, as reported in 2.2.6. Assuming that the calcium compound would

4

ideally be only calcium carbonate, we converted the Ca2+ contents into the mass equivalent of

5

CaCO3 per mass equivalent of cellulose. The values found for membranes were ~ 6.7

6

mgCC/gCellulose for samples GR30 L and GR100 L, ~ 25 mgCC/gCellulose for samples GR30 M and

7

GR100 M and ~ 48 mgCC/gCellulose for GR30 H and GR100 H (Figure 6a,b), respectively.

8 9

Figure 6 Determination of equivalent calcium carbonate content for membranes of grammage

10

100 g/m2 with low, medium and high ratio of CC/NFC (respectively GR100 L, GR100 M and

11

GR100 H) (a) and for membranes of grammage 30 g/m2 with low, medium and high ratio of

12

CC/NFC (respectively GR30 L, GR30 M and GR30 H) (b). The values were compared before

13

template removal (orange box) and after template removal (gray box) with 1 M aq. HCl for 30

14

min.

ACS Paragon Plus Environment

23

ACS Applied Materials & Interfaces

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

Page 24 of 42

1

We repeated the evaluation by complexometric titration with EDTA to assess a potential Ca2+

2

content in the neat NFC, however no calcium was detected after solvent exchange to ethanol,

3

within the resolution limit of the technique.

4

Notably, in all samples some residual CC was present; as expected titration with EDTA is more

5

sensitive to residual calcium species than FTIR spectroscopy and TGA, where the residual

6

calcium species were below the detection limit (see above). In all samples, the concentrations

7

were in the range of 2 and 7 mgCC/gCellulose. The CC concentration of samples GR100 L and

8

GR30 L remained almost unvaried upon washing. It appears that NFC is able to adsorb some

9

Ca2+ under the applied washing conditions. We assume that the calcium carbonate nanoparticles

10

arranged differently at different ratios in presence of the nanofibrillated cellulose in solvents,

11

influencing thereafter their removal.

12 13 14

3.5 Membranes properties and performance

15

3.5.1

16

Mercury intrusion results provided information on the pore size distribution and the pore stability

17

after drying. Plots of pore volume against pore size are available in the Supplementary

18

Information (Figure S5a and b): abundancy of pores below 10 nm was found in addition to pores

19

of sizes as high as 100-200 nm. No significant increase in total pore volume (areas below the

20

distribution curves) or changes in the shape of the pore size distribution were observed for

21

templated membranes with increasing CC/NFC ratios. We assume that upon repeated drying,

22

after template removal, the membrane underwent shrinking because not sufficient water to

23

ethanol exchange was achieved. In this condition, instability in the pore structure took place.

Pore size Distribution of Dry Templated Membranes

ACS Paragon Plus Environment

24

Page 25 of 42

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 Applied Materials & Interfaces

1

This result suggested that after template removal, the membranes should not be dried, but

2

directly located in the filtration apparatus at their wet state.

3

3.5.2

Water Permeance and Hydraulic Permeability of Templated Membranes

4

NFC membranes produced from water had permeance values of approximately 15

5

L/(hm2MPa) (GR 30 g/m2) and of 5 L/(hm2MPa) (GR 100 g/m2). Such values are currently

6

limiting the applicability of NFC membranes in industrial processes. Membranes prepared from

7

an ethanol-based suspension already showed a remarkably higher permeance of about 200

8

L/(hm2MPa) (GR 30 g/m2) and of 50 L/(hm2MPa) (GR 100 g/m2). This phenomenon is

9

commonly attributed to the effect of solvents which are less polar than water; in this case the

10

collapse of nanofibers on top of each other is kinetically prevented due to reduced van der Waals

11

interactions upon fast drying. This property was already observed for methanol, ethanol, acetone,

12

isopropanol and other solvents39. Membranes with different CC/NFC ratios were investigated,

13

using the washing procedure optimized in the former section 3.2 (1 M aq. HCl solution for 30

14

min). The membrane water flux was measured and the permeance calculated. For each sample

15

set, at least three replicates were tested. As shown in Figure 7a1-2, the permeance of water

16

decreased with decreasing initial CC content. However, in the first 40 min of water permeation

17

time the membrane flux was highly instable and this is attributed to the fact that the complete

18

compressibility of the membrane during use was not yet reached. To compare the performance,

19

the permeance values were considered at the steady state after 40 min. As evident from Figure 7b

20

the permeance rose with increasing initial CC content. The samples containing low amount of

21

CC (GR30 L and GR100 L) did not produce a significant improvement of the permeance

22

compared to unloaded membranes. Overall, the permeance of membranes with GR 30 g/m 2 was

23

higher than the one of membranes with GR 100 g/m2, due to a thickness effect. Thickness

ACS Paragon Plus Environment

25

ACS Applied Materials & Interfaces

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

Page 26 of 42

1

measured at the dry state, before template removal, gave values around 37-39 μm for membranes

2

of GR 30 g/m2 and values between 97 and 140 μm for membranes of GR 100 g/m2. Clearly,

3

these values tend to change after template removal, due to the swelling ability of water and also

4

because of membrane compression during filtration. A real estimation of their values in

5

operation could not be assessed. The results for GR 30 g/m2 were less reproducible than those

6

with GR 100 g/m2 because of a higher sensitivity of the thinner membranes to changes

7

introduced by the acidic washing. This can be visualized in the standard deviations of the

8

measurements in Figure 7b.

9

ACS Paragon Plus Environment

26

Page 27 of 42

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 Applied Materials & Interfaces

1

Figure 7 a1) Water permeance of membranes of grammage 30 g/m2 with low, medium and high

2

ratio of CC/NFC (respectively GR30 L, GR30 M and GR30 H) and a2) water permeance of

3

membranes of grammage 100 g/m2 with low, medium and high ratio of CC/NFC (respectively

4

GR100 L, GR100 M and GR100 H); b) comparison of average values of permeances for

5

membranes with grammage 30 g/m2 and 100 g/m2 compared to NFC membranes prepared from

6

water not subjected to acidic washing (black symbols), for membranes prepared from ethanol not

7

subjected to acidic washing (red symbols), for membranes prepared from ethanol subjected to

8

acidic washing (orange symbols), for membranes prepared from ethanol with low, medium and

9

high ratios of CC/NFC (respectively, blue, olive and green symbols); c) plot of Darcy’s law:

10

water permeability Lp corresponds to the slope of fitting lines.

11

For comparison, we also evaluated the water permeance of a commercial regenerated

12

cellulose acetate membrane. This product showed a water permeance of about 349 L/(hm2MPa)

13

(result not shown in the figures). Notably, the sample with GR30 H CC/NFC provided a water

14

permeance around 377 L/(hm2MPa), competitive with that of regenerated cellulose acetate. The

15

hydraulic permeability Lp of the membranes, a relevant parameter for defining the membrane

16

performances, was estimated taking into account Darcy’s equation plots

17

systems (Figure 7c). While all membranes with GR 30 g/m2 had Lp values in the order of ~10-11

18

m, the membranes with GR 100 g/m2 had values of ~10-12 m, as one would expect in accordance

19

with increasing thicknesses, respectively. Fitting lines did not have intercepts in zero, as one

20

would expect. This could be explained due to parallel phenomena happening when the filtration

21

has started: pore saturation with water and compressibility of the membrane while overcoming

22

the transmembrane pressure were competing during the transient state. Typical values of Lp for

23

ultrafiltration and microfiltration membranes are in the range of ~10-14 m 65.

58

for the measured

ACS Paragon Plus Environment

27

ACS Applied Materials & Interfaces

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 28 of 42

1

Apparently the main intent of enhancing the permeance of NFC membranes to values

2

comparable with those of commercially available membranes was achieved. The common

3

dissolution-regeneration processes of cellulose were avoided by application of nanofibrillar

4

cellulose. In addition, the separation ability of NFC templated membranes was qualitatively

5

evaluated.

6 7 8

3.5.3

9

Three different assessment methods were employed (see below). For this purpose polyethylene

10

glycol aqueous solutions were fed to the membranes and their rejections measured. Membranes

11

with different loading ratio CC/NFC and grammages of GR 30 g/m2 and 100 g/m2 were

12

considered and compared to neat NFC membranes (no template) produced from an ethanol

13

suspension (Figure 8a). In both grammages, similar trends were observed. First, water-based

14

membranes reach high rejection (R %) values at low molecular masses: GR 30 g/m2 gives R %

15

of 83 % at 35,000 g/mol, while GR 100 g/m2 gives R % of 90 % at 100,000 g/mol (results not

16

shown). Secondly, by increasing the loading ratio CC/NFC in ethanol prepared membranes, the

17

rejection decreases thus allowing the passage of more macromolecules through the membranes.

18

Third, thicker membranes (GR 100 g/m2) achieved better rejections compared to thinner

19

membrane (GR 30 g/m2). We conclude that, although the results were particularly satisfactory

20

for the membrane permeance with higher CC loadings (GR30 H and GR100 H), the rejection of

21

water soluble polymer resulted only in a R% of 60 %. As even large polymers can pass the

22

membranes, they could be suited as microfiltration membranes and be used for concentrating

23

fluids with suspended solids.

Filtration of Water-soluble Macromolecules, Particle Separation and Concentration of Fluids

ACS Paragon Plus Environment

28

Page 29 of 42

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 Applied Materials & Interfaces

1 2

Figure 8 Filtration tests through template membranes: a) Rejection of polyethylene glycol (PEG)

3

with different molecular masses (100,000, 900,000 and 5,000,000 g/mol) through membranes of

4

grammage 30 g/m2 (dots, interpolations with dotted lines) and through membranes of grammage

5

100 g/m2 (squares, interpolations with continuous lines); b) UV-Vis analysis of permeates after

6

concentration of orange juice through membrane of GR 100 g/m2: refractive indexes normalized

7

by the water refractive index (np/nw) and transmittances T(%) of the permeates. C1-4) Analysis of

8

permeates after filtration of water-dispersed carbon black particles through membrane of GR 30

9

g/m2: two values of the feed transmittance and value of subsequent permeates (P1, P2 and P3)

10

are plotted against wavelengths (nm).

11

Accordingly, membranes of GR 100 g/m2 were utilized to concentrate orange juice (starting

12

concentration of 12 wt%). For comparison, regenerated cellulose acetate membranes (RC) were

13

also studied. Refractive indexes normalized by the refractive index of water (1.333) were also

ACS Paragon Plus Environment

29

ACS Applied Materials & Interfaces

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 30 of 42

1

measured and indicated as np/nw. The highly concentrated feed (12 wt%) had a value of np/nw of

2

1.014. When pure water would be obtained as permeate, a value of np/nw of 1 should be expected.

3

However, the permeates through all tested membranes were above 1. The same normalized

4

refractive index of 1.012 was measured for the permeate through regenerated cellulose acetate

5

membrane (RC), for unloaded membranes (EtOH GR100) and for templated membranes with

6

low loading (GR100 L). Membranes with increasing loading (GR100 M and GR100 H) had

7

values of np/nw slightly lower, in the order of 1.011. The permeates were also analyzed by UV-

8

Vis measurements to determine the transmittance T (%) (Figure 8b). While the feed had a

9

transmittance of almost zero, as expected from its high turbidity, the permeates showed different

10

values for RC and templated membranes. Combining the UV-Vis results with the refractive

11

indexes values, we qualitatively conclude that there was permeation of some compounds which

12

passed through the membrane, such as sugars and chromophore molecules, while solids where

13

mostly rejected. A visual evaluation of the suspended elements of orange juice is provided in

14

Supplementary Information, Figure S6: fibers with lengths above ca. 100 μm and clusters of

15

particles of a few μm were found. In addition, adsorption of chromophore compounds in the

16

NFC membranes was also observed.

17

Membranes with lower grammage were also tested for their ability to remove small particles. As

18

convenient example, carbon black nanoparticles and microparticles (primary particles diameter >

19

100 nm and secondary particles with diameters up to 100 μm) were dispersed in water at a

20

starting concentration of 0.05 g/L and filtered through templated membranes of GR 30 g/m2 (the

21

original particle size distribution of carbon black is presented in the supplementary information,

22

section S6.1, Figures S7 and S8). The transmittance of the feed and of three subsequent

23

permeates (P1, P2 and P3) was analyzed with UV-Vis spectroscopy. The transmittance of the

ACS Paragon Plus Environment

30

Page 31 of 42

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 Applied Materials & Interfaces

1

feed was only 6-11 %, the value drastically increased already at the first permeate, independent

2

from the membrane type. A clarification of 80 % - 100 % was achieved (Figure 8c1-4), i.e. for

3

some membranes the carbon black was almost completely removed. Hence, it appeared that the

4

membranes are suitable for filtration of nano- and microparticles (particle size > 100 nm). In

5

Figure 9a the orange juice feed, permeates for each membrane type and the membranes

6

themselves are shown. The permeate for RC had a darker color compared with the one obtained

7

from the templated membranes, in agreement with its higher refractive index aforementioned. A

8

picture of the starting carbon black feed and the permeates (P1) is shown in Figure 9b: from right

9

to left: membranes covered by carbon black particles removed during filtration and vials

10

containing feed and permeates P1 for each system.

11 12

Figure 9 From left to right: a) Orange juice at 12 % dry matter content, permeates collected after

13

4 h for regenerated cellulose acetate membrane (RC) and templated membranes with GR 100

14

g/m2 of low (L), medium (M) and high (H) CC loading. b) Carbon Black feed (black

ACS Paragon Plus Environment

31

ACS Applied Materials & Interfaces

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 32 of 42

1

suspension), permeates P1 through each membranes with GR 30 g/m2 of low (L), medium (M)

2

and high (H) CC loading covered by carbon black particles removed after filtration.

3

We conclude that since carbon black is uncharged, the main rejection mechanism is due to size-

4

exclusion of the retained suspended particles. However, in the case of orange juice (its exact

5

composition is unknown), adsorption mechanisms might also occur involving the hydroxyl

6

groups of cellulose which might interact with polar groups of orange juice components or by

7

hydrophobic interactions. Certainly more detailed investigations of the separation efficiency of

8

the membranes, on the permeates compositions, on the fouling behavior and change of

9

selectivity over time are needed. In the case of hydrophobic interactions, preliminary rejection

10

tests with proteins (viz. bovin serum albumin and immunoglobin G) demonstrated some

11

interactions, which need further investigations. However, these aspects will be in the scope of

12

follow-up studies.

13

4

CONCLUSIONS

14

Porous membranes could be produced by a templating approach, based on incorporation of

15

calcium compound (CC) particles such as calcium carbonate, which were removed by acidic

16

washing of CC/NFC films prepared from ethanolic suspensions. The ratio between CC and NFC

17

could be controlled readily. A crucial point was the optimization of the CC removal procedure by

18

acidic washing. By drying the membranes from ethanol, already a substantial increase of the

19

permeance of such membranes was found. In addition, an increase in permeance was noticed by

20

application of CC particles as template. Overall, the water permeability could be enhanced by 2 –

21

3 times, when compared to membranes prepared from water, leading to values in the range of

22

those obtained for commercial cellulose membrane systems. High liquid fluxes and permeation

ACS Paragon Plus Environment

32

Page 33 of 42

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 Applied Materials & Interfaces

1

of polymer macromolecules were observed, but the retention of small solid particles was shown

2

as example. Accordingly, concentration of orange juice with suspended solids was envisaged.

3

Given also the organic nature of the cellulose, one could propose for our membranes applications

4

in the food and beverage industries, such as removal of bacteria and spores, filtration of extra-

5

virgin olive-oil,

6

concentration of solutions in dairy and sugar industry

7

assess such functions. In addition, after template removal, specific chemical functionalization

8

could be imparted to NFC, empowering specific compatibilities with different chemical

9

environments and aiming at fouling reduction69. While we used particles as hard template, other

10

solid soft and hard templates with different particle aspect ratios might be considered in order to

11

modify the shape of pores after template removal. The proposed strategy is based on the use of

12

sustainable materials, such as cellulose, ethanol and calcium carbonate. Moreover, NFC

13

membranes are biodegradable and can be produced with a green and cost-effective process, in

14

line with the materials design principle of a cradle-to-cradle approach70 and low environmental

15

impact at end-of-life disposal.

low-temperature clarification and concentration of juices and beer, 66–68

. Further investigations are needed to

16 17

ASSOCIATED CONTENT

18

Section S1.1 CC particles: SEM, FTIR, TGA, XRD; section S2.1 Preparation of templated

19

membranes with different grammages; section S3.1 Calibration curve for the determination of

20

calcium carbonate by EDTA back-titration; section S4.1 Carbon Black particle size distribution.

21

Supporting material is available free of charge via the Internet at http://pubs.acs.org.

22

AUTHOR INFORMATION

ACS Paragon Plus Environment

33

ACS Applied Materials & Interfaces

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 34 of 42

1

Corresponding Authors

2

* [email protected], +41587656172

3

* [email protected], +41587654115

4

Author Contributions

5

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

6

to the final version of the manuscript.

7

Funding Sources

8

Empa is kindly acknowledged for financial support.

9

ACKNOWLEDGMENT

10

We kindly acknowledge EMPA colleagues Dr. Balogh for X-ray analysis, B. Fisher for TGA

11

measurements, E. Strub and Anja Huch for SEM imaging, Dr. Erni for the TEM-SAED

12

evaluations.

13 14

REFERENCES

15

(1)

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

16

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

17

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

18

(2)

Qing, Y.; Sabo, R.; Zhu, J. Y.; Agarwal, U.; Cai, Z. Y.; Wu, Y. Q. A Comparative Study

19

of Cellulose Nanofibrils Disintegrated via Multiple Processing Approaches. Carbohydr.

20

Polym. 2013, 97 (1), 226–234.

21 22

(3)

Siró, I.; Plackett, D. Microfibrillated Cellulose and New Nanocomposite Materials: A Review. Cellulose 2010, 17 (3), 459–494.

ACS Paragon Plus Environment

34

Page 35 of 42

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

1

ACS Applied Materials & Interfaces

(4)

Tejado, A.; Alam, M. N.; Antal, M.; Yang, H.; Ven, T. G. M. Energy Requirements for

2

the Disintegration of Cellulose Fibers into Cellulose Nanofibers. Cellulose 2012, 19 (3),

3

831–842.

4

(5)

Turbak, A. F.; Snyder, F. W.; Sandberg, K. R. Microfibrillated Cellulose, a New Cellulose

5

Product: Properties, Uses, and Commercial Potential. J. Appl. Polym. Sci. Appl. Polym.

6

Symp. 1983, 37, 815–827.

7

(6)

Ho, T. T. T.; Ko, Y. S.; Zimmermann, T.; Geiger, T.; Caseri, W. Processing and

8

Characterization of Nanofibrillated Cellulose/layered Silicate Systems. J. Mater. Sci.

9

2012, 47 (10), 4370–4382.

10

(7)

Aulin, C.; Gällstedt, M.; Lindström, T.; Gallstedt, M.; Lindstrom, T. Oxygen and Oil

11

Barrier Properties of Microfibrillated Cellulose Films and Coatings. Cellulose 2010, 17

12

(3), 559–574.

13

(8)

Kelley L. Spence, Richard A. Venditti, Orlando J. Rojas, JOel J. Pawlak, M. A. H. Water

14

Vapor Barrier Properties of Coated and Filled Microfibrillated Cellulose Composite

15

Films. BioResources 2011, 6, 4370–4388.

16

(9)

Lavoine, N.; Desloges, I.; Dufresne, A.; Bras, J. Microfibrillated Cellulose - Its Barrier

17

Properties and Applications in Cellulosic Materials: A Review. Carbohydr. Polym. 2012,

18

90 (2), 735–764.

19

(10)

20 21

Syverud, K.; Stenius, P. Strength and Barrier Properties of MFC Films. Cellulose 2009, 16 (1), 75–85.

(11)

Wu, C. N.; Saito, T.; Fujisawa, S.; Fukuzumi, H.; Isogai, A. Ultrastrong and High Gas-

22

Barrier Nanocellulose/Clay-Layered Composites. Biomacromolecules 2012, 13 (6), 1927–

23

1932.

24

(12)

Zhang, Z.; Sèbe, G.; Rentsch, D.; Zimmermann, T.; Tingaut, P. Ultralightweight and

25

Flexible Silylated Nanocellulose Sponges for the Selective Removal of Oil from Water.

26

Chem. Mater. 2014, 26 (8), 2659–2668.

27

(13)

Jin, H.; Kettunen, M.; Laiho, A.; Pynnonen, H.; Paltakari, J.; Marmur, A.; Ikkala, O.; Ras,

ACS Paragon Plus Environment

35

ACS Applied Materials & Interfaces

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 36 of 42

1

R. H. A. Superhydrophobic and Superoleophobic Nanocellulose Aerogel Membranes as

2

Bioinspired Cargo Carriers on Water and Oil. Langmuir 2011, 27 (5), 1930–1934.

3

(14)

Gebald, C.; Wurzbacher, J. A.; Tingaut, P.; Zimmermann, T.; Steinfeld, A. Amine-Based

4

Nanofibrillated Cellulose as Adsorbent for CO(2) Capture from Air. Environ. Sci.

5

Technol. 2011, 45 (20), 9101–9108.

6

(15)

Mautner, A.; Lee, K. Y.; Lahtinen, P.; Hakalahti, M.; Tammelin, T.; Li, K.; Bismarck, A.

7

Nanopapers for Organic Solvent Nanofiltration. Chem. Commun. (Cambridge, U. K.)

8

2014, 50 (43), 5778–5781.

9

(16)

Mautner, A.; Lee, K.-Y. Y.; Tammelin, T.; Mathew, A. P.; Nedoma, A. J.; Li, K.;

10

Bismarck, A. Cellulose Nanopapers as Tight Aqueous Ultra-Filtration Membranes. React.

11

Funct. Polym. 2015, 86, 209–214.

12

(17)

Sehaqui, H.; de Larraya, U. P.; Liu, P.; Pfenninger, N.; Mathew, A. P.; Zimmermann, T.;

13

Tingaut, P. Enhancing Adsorption of Heavy Metal Ions onto Biobased Nanofibers from

14

Waste Pulp Residues for Application in Wastewater Treatment. Cellulose 2014, 21 (4),

15

2831–2844.

16

(18)

Sehaqui, H.; Mautner, A.; de Larraya, U.; Pfenninger, N.; Tingaut, P.; Zimmermann, T.

17

Cationic Cellulose Nanofibers from Waste Pulp Residues and Their Nitrate, Fluoride,

18

Sulphate and Phosphate Adsorption Properties. Carbohydr. Polym. 2016, 135, 334–340.

19

(19)

Ferraz, N.; Leschinskaya, A.; Toomadj, F.; Fellström, B.; Strømme, M.; Mihranyan, A.

20

Membrane Characterization and Solute Diffusion in Porous Composite Nanocellulose

21

Membranes for Hemodialysis. Cellulose 2013, 20 (6), 2959–2970.

22

(20)

Razaq, A.; Nyström, G.; Strømme, M.; Mihranyan, A.; Nyholm, L. High-Capacity

23

Conductive Nanocellulose Paper Sheets for Electrochemically Controlled Extraction of

24

DNA Oligomers. PLoS One 2011, 6 (12), e29243.

25

(21)

Metreveli, G.; Wågberg, L.; Emmoth, E.; Belák, S.; Strømme, M.; Mihranyan, A. A Size-

26

Exclusion Nanocellulose Filter Paper for Virus Removal. Adv. Healthc. Mater. 2014, 3

27

(10), 1546–1550.

ACS Paragon Plus Environment

36

Page 37 of 42

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

1

ACS Applied Materials & Interfaces

(22)

2 3

Treatment Technologies. Environ. Sci. Technol. 2015, 49 (9), 5277–5287. (23)

4 5

(24)

(25)

Ma, H.; Burger, C.; Hsiao, B. S.; Chu, B. Ultrafine Polysaccharide Nanofibrous Membranes for Water Purification. Biomacromolecules 2011, 12 (4), 970–976.

(26)

10 11

Ma, H.; Burger, C.; Hsiao, B. S.; Chu, B. Ultra-Fine Cellulose Nanofibers: New NanoScale Materials for Water Purification. J. Mater. Chem. 2011, 21 (21), 7507–7510.

8 9

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

6 7

Carpenter, A. W.; de Lannoy, C. F.; Wiesner, M. R. Cellulose Nanomaterials in Water

Ma, H.; Burger, C.; Hsiao, B. S.; Chu, B. Nanofibrous Microfiltration Membrane Based on Cellulose Nanowhiskers. Biomacromolecules 2012, 13 (1), 180–186.

(27)

Ma, H.; Burger, C.; Hsiao, B. S.; Chu, B. Fabrication and Characterization of Cellulose

12

Nanofiber Based Thin-Film Nanofibrous Composite Membranes. J. Memb. Sci. 2014, 454,

13

272–282.

14

(28)

15 16

Wang, X.; Fang, D.; Hsiao, B. S.; Chu, B. Nanofiltration Membranes Based on Thin-Film Nanofibrous Composites. J. Memb. Sci. 2014, 469, 188–197.

(29)

Orsolini, P.; Michen, B.; Huch, A.; Tingaut, P.; Caseri, W. R.; Zimmermann, T.

17

Characterization of Pores in Dense Nanopapers and Nanofibrillated Cellulose Membranes:

18

A Critical Assessment of Established Methods. ACS Appl. Mater. Interfaces 2015, 7 (46),

19

25884–25897.

20

(30)

Karim, Z.; Claudpierre, S.; Grahn, M.; Oksman, K.; Mathew, A. P. Nanocellulose Based

21

Functional Membranes for Water Cleaning: Tailoring of Mechanical Properties, Porosity

22

and Metal Ion Capture. J. Memb. Sci. 2016, 514, 418–428.

23

(31)

Mautner, A.; Maples, H. A.; Sehaqui, H.; Zimmermann, T.; Perez de Larraya, U.;

24

Mathew, A. P.; Lai, C. Y.; Li, K.; Bismarck, A. Nitrate Removal from Water Using a

25

Nanopaper Ion-Exchanger. Environ. Sci. Water Res. Technol. 2016, 2 (1), 117–124.

26

(32)

Sehaqui, H.; Perez de Larraya, U.; Tingaut, P.; Zimmermann, T. Humic Acid Adsorption

ACS Paragon Plus Environment

37

ACS Applied Materials & Interfaces

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 38 of 42

1

onto Cationic Cellulose Nanofibers for Bioinspired Removal of Copper(II) and a

2

Positively Charged Dye. Soft Matter 2015, 11 (26), 5294–5300.

3

(33)

4 5

De Gruyter: Berlin, 2012. (34)

6 7

Rodionova, G.; Hoff, B.; Lenes, M.; Eriksen, Ø.; Gregersen, Ø. Gas-Phase Esterification of Microfibrillated Cellulose (MFC) Films. Cellulose 2013, 20 (3), 1167–1174.

(35)

8 9

Dufresne, A. Nanocellulose from Nature to High Performance Tailored Materials; Walter

Andresen, M.; Johansson, L.-S.; Tanem, B. S.; Stenius, P. Properties and Characterization of Hydrophobized Microfibrillated Cellulose. Cellulose 2006, 13 (6), 665–677.

(36)

Tingaut, P.; Zimmermann, T.; Lopez-Suevos, F. Synthesis and Characterization of

10

Bionanocomposites with Tunable Properties from Poly(lactic Acid) and Acetylated

11

Microfibrillated Cellulose. Biomacromolecules 2010, 11 (2), 454–464.

12

(37)

Berlioz, S.; Molina-Boisseau, S.; Nishiyama, Y.; Heux, L. Gas-Phase Surface

13

Esterification of Cellulose Microfibrils and Whiskers. Biomacromolecules 2009, 10 (8),

14

2144–2151.

15

(38)

16 17

Sehaqui, H.; Zimmermann, T.; Tingaut, P. Hydrophobic Cellulose Nanopaper through a Mild Esterification Procedure. Cellulose 2014, 21 (1), 367–382.

(39)

Svensson, A.; Larsson, P. T.; Salazar-Alvarez, G.; Wagberg, L. Preparation of Dry Ultra-

18

Porous Cellulosic Fibres: Characterization and Possible Initial Uses. Carbohydr. Polym.

19

2013, 92 (1), 775–783.

20

(40)

Peng, Y.; Gardner, D. J.; Han, Y.; Kiziltas, A.; Cai, Z.; Tshabalala, M. A. Influence of

21

Drying Method on the Material Properties of Nanocellulose I: Thermostability and

22

Crystallinity. Cellulose 2013, 20 (5), 2379–2392.

23

(41)

24 25 26

Peng, Y.; Gardner, D. J.; Han, Y. Drying Cellulose Nanofibrils: In Search of a Suitable Method. Cellulose 2011, 19 (1), 91–102.

(42)

Henriksson, M.; Berglund, L. A.; Isaksson, P.; Lindstrom, T.; Nishino, T. Cellulose Nanopaper Structures of High Toughness. Biomacromolecules 2008, 9 (6), 1579–1585.

ACS Paragon Plus Environment

38

Page 39 of 42

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

1

ACS Applied Materials & Interfaces

(43)

Petkovich, N. D.; Stein, A. Controlling Macro- and Mesostructures with Hierarchical

2

Porosity through Combined Hard and Soft Templating. Chem. Soc. Rev. 2013, 3721–

3

3739.

4

(44)

5 6

1997, 389 (6654), 948–951. (45)

7 8

Caruso, F.; Caruso, R. A.; Mohwald, H. Nonoengeneering of Inorganic and Hybrid Hollow Spheres by Colloidal Templating. Science 1998, 282, 1111–1114.

(46)

9 10

Pine, D. J.; Imhof, A. Ordered Macroporous Materials by Emulsion Templating. Nature

Pulko, I.; Krajnc, P. High Internal Phase Emulsion Templating - A Path to Hierarchically Porous Functional Polymers. Macromol. Rapid Commun. 2012, 33 (20), 1731–1746.

(47)

Wilke, A.; Weber, J. Hierarchical Nanoporous Melamine Resin Sponges with Tunable

11

Porosity—porosity Analysis and CO2 Sorption Properties. J. Mater. Chem. 2011, 21 (14),

12

5226.

13

(48)

14 15

Nanostructured Organic Material. Chem. Mater. 2008, 20 (3), 738–755. (49)

16 17

Thomas, A.; Goettmann, F.; Antonietti, M. Hard Templates for Soft Materials: Creating

Chen, S. Q.; Chen, J.; Du, R. X. Fabrication of Water Treatment Membrane Using Templating Method - A Critical Review. Adv. Mater. Res. 2011, 343–344, 130–141.

(50)

Kellenberger, C. R.; Luechinger, N. A.; Lamprou, A.; Rossier, M.; Grass, R. N.; Stark, W.

18

J. Soluble Nanoparticles as Removable Pore Templates for the Preparation of Polymer

19

Ultrafiltration Membranes. J. Memb. Sci. 2012, 387–388 (1), 76–82.

20

(51)

Kellenberger, C. R.; Pfleiderer, F. C.; Raso, R. a.; Burri, C. H.; Schumacher, C. M.; Grass,

21

R. N.; Stark, W. J. Limestone Nanoparticles as Nanopore Templates in Polymer

22

Membranes: Narrow Pore Size Distribution and Use as Self-Wetting Dialysis Membranes.

23

RSC Adv. 2014, 4 (106), 61420–61426.

24

(52)

Kellenberger, C. R.; Hess, S. C.; Schumacher, C. M.; Loepfe, M.; Nussbaumer, J. E.;

25

Grass, R. N.; Stark, W. J. Roll-to-Roll Preparation of Mesoporous Membranes by

26

Nanoparticle Template Removal. Ind. Eng. Chem. Res. 2014, 53 (22), 9214–9220.

ACS Paragon Plus Environment

39

ACS Applied Materials & Interfaces

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

1

(53)

Page 40 of 42

Chen, S.; Cölfen, H.; Antonietti, M.; Yu, S. Ethanol Assisted Synthesis of Pure and Stable

2

Amorphous Calcium Carbonate Nanoparticles. Chem. Commun. (Cambridge, U. K.) 2013,

3

49 (83), 9564–9566.

4

(54)

5 6

Analysis. Nat. Methods 2012, 9 (7), 671–675. (55)

7 8

DIN 54372, Prüfung von Zellstoff Und Papier. Bestimmung Des Calcium- Und Magnesiumgehaltes. Titrimetrische Bestimmung. p 54–56 (1988–04).

(56)

9 10

Schneider, C. a; Rasband, W. S.; Eliceiri, K. W. NIH Image to ImageJ: 25 Years of Image

Allemann, S. Die Nicht-Wässrige Entsäuerung von Transparentpapier, Master Thesis, Hochschule der Künste Bern, 2012.

(57)

Segal, L.; Creely, J. J.; Martin, a. E.; Conrad, C. M. An Empirical Method for Estimating

11

the Degree of Crystallinity of Native Cellulose Using the X-Ray Diffractometer. Text.

12

Res. J. 1959, 29 (10), 786–794.

13

(58)

14 15

Sci. Eng.; 2010; Vol. 1, pp 311–335. (59)

16 17

Schärtl, W. Light Scattering from Polymer Solutions and Nanoparticle Dispersions; Springer Berlin Heidelberg, 2013; Vol. 53.

(60)

18 19

Causserand, C.; Aimar, P. Characterization of Filtration Membranes. In Compr. Membr.

Istchenko, R.; No, O. ISO 22412 Paticle Size Analysis -- Dynamic Light Scattering (DLS). 2006, 2005 (Dld), 22674.

(61)

Mohammadkazemi, F.; Faria, M.; Cordeiro, N. In Situ Biosynthesis of Bacterial

20

Nanocellulose-CaCO3 Hybrid Bionanocomposite: One-Step Process. Mater. Sci. Eng. C

21

2016, 65, 393–399.

22

(62)

23 24

Growth 1989, 98 (3), 504–510. (63)

25 26

Brecevic, L.; Nielsen, A. E. Solubility of Amorphous Calcium Carbonate. J. Cryst.

Battista, O. a. Hydrolysis and Crystallization of Cellulose. Ind. Eng. Chem. 1950, 42 (3), 502–507.

(64)

Dash, S.; Kamruddin, M.; Ajikumar, P. K.; Tyagi, A. K.; Raj, B. Nanocrystalline and

ACS Paragon Plus Environment

40

Page 41 of 42

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 Applied Materials & Interfaces

1

Metastable Phase Formation in Vacuum Thermal Decomposition of Calcium Carbonate.

2

Thermochim. Acta 2000, 363 (1–2), 129–135.

3

(65)

Huisman, I. H.; Dutré, B.; Persson, K. M.; Trägårdh, G. Water Permeability in

4

Ultrafiltration and Microfiltration: Viscous and Electroviscous Effects. Desalination 1997,

5

113 (1), 95–103.

6

(66)

7 8

Springer-Verlag, 2014. (67)

9

2013, 4 (9), 1000269. (68)

12 13

Trägardh, G.; Gekas, V. Membrane Technology in the Sugar Industry. Desalination 1988, 69, 9–17.

(69)

14 15

Marella C, K Muthukumarappan, L. E. M. Application of Membrane Separation Technology for Developing Novel Dairy Food Ingredients. J. Food Process. Technol.

10 11

Ahmad, S. Food Processing: Strategies for Quality Assessment; SK, Mushir, S. A., Ed.;

Rana, D.; Matsuura, T. Surface Modifications for Antifouling Membranes. Chem. Rev. 2010, 110 (4), 2448–2471.

(70)

McDonough, W.; Braungart, M.; Anastas, P. T.; Zimmerman, J. B. Peer Reviewed:

16

Applying the Principles of Green Engineering to Cradle-to-Cradle Design. Environ. Sci.

17

Technol. 2003, 37 (23), 434A–441A.

18

19

ACS Paragon Plus Environment

41

ACS Applied Materials & Interfaces

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

Table of content 468x187mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 42 of 42