1. business and industrial
2. chemicals industry
3. plastics and polymers

PREPARATION AND CHARACTERIZATION OF LIGNIN NANOFIBRE-BASED
MATERIALS OBTAINED BY ELECTROSTATIC SPINNING
by
James Ian Dallmeyer
A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF
THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
in
THE FACULTY OF GRADUATE STUDIES
(Forestry)
THE UNIVERSITY OF BRITISH COLUMBIA
(Vancouver)
March 2013
© James Ian Dallmeyer, 2013
Abstract
Electrostatic spinning was investigated as a means to generate nanofibre-based
materials from lignin. Kraft, organosolv, lignosulfonate, and pyrolytic lignins were found to
be prone to electrospray, resulting in the formation of droplets instead of fibres upon
electrical charging of their solutions in most cases. It was observed that the addition of a
small amount of poly(ethylene oxide) (PEO) to the spinning solution was an effective
strategy to promote the formation of uniform fibres. Studies on the shear and elongational
rheology of the spinning solutions were conducted to understand the mechanism underlying
the improved process stability that resulted from the addition of PEO to lignin solutions. It
was found that the shear rheology was changed to a small extent upon addition of PEO to the
spinning solution, while studies using capillary breakup extensional rheometry revealed that
PEO addition induced non-Newtonian, strain hardening behaviour to lignin solutions, which
was undetectable in shear rheology studies. The concentration of lignin in solution,
concentration of PEO, and molecular weight of PEO were shown to influence the
elongational fluid properties, which displayed a strong correlation with the fibre diameter.
Once the rheology of the spinning solution was characterized, attention was focused on
oxidative thermostabilization and carbonization of electrospun nonwoven fabrics. It was
found that incorporating different amounts of Kraft lignin fractions in electrospun fibres
allowed the preparation of interesting material morphologies depending on the thermal
softening characteristics of Kraft lignin fractions. Two interesting types of materials were
prepared by controlling the composition of lignin fabrics and processing parameters. The first
was a novel stimuli-responsive film material with a reversible ability to change shape in
response to moisture. The second type was an interconnected carbon nanofibre network
which displayed interesting mechanical, surface, and electrical properties. The structure and
properties of Kraft lignin fractions were investigated by thermo-rheological analysis, nuclear
magnetic resonance spectroscopy, gel permeation chromatography, and light scattering to
understand the role of lignin structure in determining the properties of thermostabilized and
carbonized lignin fabrics. The electrospun fabrics were also characterized by atomic force
microscopy, microtensile testing, nitrogen adsorption, X-ray diffraction, and Raman
spectroscopy.
i
Preface
A version of chapter 3 has been published in the Journal of Wood Chemistry and
Technology (Dallmeyer, I., Ko, F., Kadla, J.F. J. Wood Chem. And Tech. 2010, 30, 4, 315329). I conducted all of the experiments and data analysis described and wrote the
manuscript, and Professors Frank Ko and John Kadla suggested corrections to improve the
manuscript.
A version of chapter 4 has been submitted for publication (Dallmeyer, I., Ko, F.,
Kadla, J.F. submitted for review in 2012) I conducted all of the experiments described in
chapter 4 and wrote the manuscript, and Professors Frank Ko and John Kadla advised me
with corrections to improve the manuscript. The contribution of Professor Savvas
Hatzikiriakos, who also advised me with suggestions to improve the manuscript, is also
acknowledged.
Chapter 5 includes some data that was collected by Dr. Sudip Chowdhury. He was
responsible for conducting thermorheological analysis of Kraft lignin fractions and wrote
portions of the sections describing the data (section 5.6). I wrote everything else in Chapter 5
and conducted all of the other experimental work.
Mr. Li Ting Lin, Ms. Yingjie (Phoebe) Li, and Ms. Nai-Yu Teng helped conduct
some of the experimental work described in Chapter 6. Li Ting conducted some of the tensile
testing (section 6.3) and collected some of the Raman spectroscopy data (section 6.6).
Phoebe Li prepared electrospun polyacrylonitrile (PAN) fibres for comparison with fibres
prepared from lignin. Nai-Yu Teng collected the x-ray diffraction patterns shown in the
Appendix. I conducted all of the rest of the experimental work and wrote the entire
manuscript.
ii
Table of Contents
Abstract ..................................................................................................................................... i
Preface ...................................................................................................................................... ii
Table of Contents ................................................................................................................... iii
List of Tables .......................................................................................................................... vi
List of Figures ........................................................................................................................ vii
List of Symbols and Abbreviations ...................................................................................... xi
Acknowledgements .............................................................................................................. xiii
Dedication ............................................................................................................................. xiv
Chapter 1. Introduction ......................................................................................................... 1
1.1
Motivation for development of advanced lignin-based nanofibre materials:........................ 1
1.2
Lignin biosynthesis and structure: ........................................................................................ 5
1.3
Isolation of lignin by chemical pulping: ............................................................................. 11
1.4
Electrospinning: .................................................................................................................. 16
1.5
Stimuli-responsive shape memory materials from lignin: .................................................. 22
1.6
Production of carbon fibres from lignin: ............................................................................. 26
Chapter 2. Materials and experimental methods............................................................... 31
2.1
Materials: ............................................................................................................................ 31
2.2
Fractionation of SKL by sequential solvent extraction: ...................................................... 32
2.3
Lignin characterization: ...................................................................................................... 32
2.3.1
Thermorheological analysis of F4 and F1-3:..................................................................... 32
2.3.2
Acetylation of lignin: ...................................................................................................... 33
2.3.3
Characterization of lignin by nuclear magnetic resonance (NMR) spectroscopy: ......... 33
2.3.4
Characterization of lignin molecular weight distribution by gel permeation
chromatography and multi-angle laser light scattering (GPC-MALLS): .................................... 34
2.4
Preparation of lignin-PEO solutions for electrospinning: ................................................... 34
2.5
Electrospinning: .................................................................................................................. 35
2.6
Rheological characterization of electrospinning solutions: ................................................ 36
2.6.1
Steady shear viscosity measurements: ............................................................................ 36
iii
2.6.2
Small amplitude oscillatory shear rheometry: ................................................................ 36
2.6.3
Capillary breakup extensional rheometry: ...................................................................... 36
2.7
Thermostabilization of electrospun fabrics: ........................................................................ 37
2.8
Carbonization of electrospun fabrics: ................................................................................. 38
2.9
Characterization of electrospun fabrics:.............................................................................. 39
2.9.1
Optical and scanning electron microscopy of fibres obtained by electrospinning: ........ 39
2.9.2
Atomic force microscopy (AFM) of thermostabilized electrospun materials: ............... 39
2.9.3
Characterization of mechanical properties of electrospun fabrics: ................................. 40
2.9.4
Electrical conductivity of carbonized materials: ............................................................ 40
2.9.5
BET surface area measurements on carbonized electrospun materials: ......................... 41
2.9.6
Characterization of carbonized materials by Raman spectroscopy: ............................... 41
Chapter 3. Electrospinning of technical lignins for the production of fibrous networks
................................................................................................................................................. 42
3.1
Introduction: ........................................................................................................................ 42
3.2
Electrospinning of technical lignin solutions without PEO: ............................................... 42
3.3
Electrospinning of lignin with addition of PEO:................................................................. 44
3.4
Effect of shear viscosity on fibre formation and diameter: ................................................. 48
3.5
Conclusion: ......................................................................................................................... 52
Chapter 4. Effect of elongational rheology on electrospinning of softwood Kraft lignin54
4.1
Introduction: ........................................................................................................................ 54
4.2
Solutions: ............................................................................................................................ 56
4.3
Dynamic shear rheometry: .................................................................................................. 56
4.4
Elongational rheometry by CaBER: ................................................................................... 58
4.5
Ability of F4 and F4/PEO solutions to form fibres during electrospinning: ........................ 64
4.6
Correlation of relaxation time with fibre diameter: ............................................................ 67
4.7
Conclusion: ......................................................................................................................... 69
Chapter 5. Preparation of moisture-responsive lignin-based materials .......................... 70
5.1
Introduction: ........................................................................................................................ 70
5.2
Electrospinning of lignin fractions F4 and F1-3: ................................................................... 70
5.3
Thermostabilization of nonwoven fabrics:.......................................................................... 73
5.4
Moisture-responsiveness, shape change, and shape recovery: ............................................ 76
5.5
AFM imaging of moisture-responsive films: ...................................................................... 79
5.6
Dynamic rheology of lignin fractions: ................................................................................ 83
iv
5.7
NMR characterization of lignin fractions: .......................................................................... 87
5.8
Characterization of molecular weight by GPC-MALLS: ................................................... 94
5.9
Conclusion: ......................................................................................................................... 97
Chapter 6. Preparation and characterization of interconnected lignin-based carbon
nanofibre materials ............................................................................................................... 98
6.1
Introduction: ........................................................................................................................ 98
6.2
Carbonization of thermostabilized electrospun nonwovens: .............................................. 98
6.3
Effect of inter-fibre bonding on mechanical properties: ................................................... 101
6.4
Electrical conductivity of carbonized fabrics:................................................................... 104
6.5
BET surface area of carbonized fibres: ............................................................................. 106
6.6
Raman spectroscopy of carbonized fibres: ....................................................................... 106
6.7
Conclusion: ....................................................................................................................... 112
Chapter 7. Concluding Remarks ....................................................................................... 114
References………………………………………………………………………………….122
Appendix………………………………………………………………………………...…151
v
List of Tables
Table 1.1: Pulping processes and reactive species.................................................................. 12
Table 3.2: Mean fibre diameters + standard deviation (n = 200) for 99/1 lignin/PEO fibres
obtained from different technical lignins ................................................................................ 47
Table 4.1: Fluid compositions, viscosity , relaxation time , surface tension , and
corresponding fibre diameters. - = too small to measure, *= incomplete fibre solidification
during spinning, x = no fibres formed, n/a = not applicable. n represents the number of
samples used to obtain solution properties. ............................................................................ 55
Table 5.1: Spinning solution compositions, thermostabilization heating rates, and resulting
morphology after heating electrospun F4/F1-3 fibres ............................................................... 73
Table 5.2: Integration of quantitative 13C-NMR spectra of F1-3 and F4. The area of the
aromatic region (162-102 ppm) was set to 600. The reported values are therefore reported as
quantities per 100 aromatic rings. Ar = aromatic, Alk = alkyl. .............................................. 93
Table 6.1: Diameters of as-spun, thermostabilized, and carbonized lignin NBF materials.. 101
Table 6.2: Mechanical properties of bonded and non-bonded fabrics before and after
carbonization at 1000oC. Values are expressed as mean + one standard deviation based on
measurements on N samples. ................................................................................................ 101
vi
List of Figures
Figure 1.1: Lignin monomers (a,b,c) and subunits (d,e,f) (a) p-coumaryl alcohol (b)
coniferyl alcohol (c) sinapyl alcohol (d) p-hydroxyphenyl (H) subunit (e) guaiacyl (G)
subunit (f) syringyl (S) subunit ................................................................................................. 6
Figure 1.2: Characteristic linkages in lignin (a) -aryl ether (-O-4) (b) phenylcoumaran (5) (c) resinol (-) (d) biphenyl (5-5’) (e) biphenyl ether (4-O-5) (f) dibenzodioxocin (-O4/-O-4/5-5’) ............................................................................................................................ 8
Figure 1.3: Schematic illustration of electrospinning process ............................................... 17
Figure 1.4: Scanning electron micrograph of electrospun PAN nanofibre nonwoven fabric.18
Figure 3.1: SEM images of SKL solutions electrospun at 40 wt% (a), and 50 wt% (b). Scale
bar = 100 m ........................................................................................................................... 43
Figure 3.2: SEM images of 95/5 and 99/1 SKL/PEO fibres electrospun from solutions at
different concentrations. (a) 95/5, 20 wt%, (b) 95/5, 25 wt%, (c) 95/5, 30 wt%, (d) 99/1, 30
wt%, (e) 99/1, 35 wt%, (f) 99/1, 40 wt%. Scale bar = 20 m (2000X magnification). .......... 45
Figure 3.3: SEM images of lignin fibres electrospun from 99/1, lignin/PEO solutions using
different technical lignins. (a) HKL 40 wt%; (b) PL 40 wt%; (c) SL 30 wt%; (d) SOL 50
wt%; (e) HOL 40%; (f) LS 30 wt%. All scale bars = 20 m (2000X magnification). ........... 46
Figure 3.4: Plot of fibre diameter vs. concentration for the 99/1 SKL/PEO system. Diameters
are reported as mean + standard deviation, based on 100 fibers for each solution and n=2
solutions prepared at each concentration. ............................................................................... 48
Figure 3.5: Plot of specific viscosity vs. concentration for SKL and SKL/PEO systems. .... 49
Figure 4.1: (a) Stress sweep and (b) frequency sweep data for F4 and F4/PEO solutions with
F4 concentration = 40 wt%...................................................................................................... 56
Figure 4.2: Representative thinning profiles of F4 and F4/PEO solutions. ............................ 58
Figure 4.3: (a) Transient elongational viscosity of F4/PEO solutions with F4 conc. = 40
wt%. (b) Semi-log plot of thinning profiles of F4/PEO solutions with F4 conc. = 40 wt%. (c)
Region of CaBER data close to filament breakup showing data scatter at small filament
diameters.. ............................................................................................................................... 61
Figure 4.4: SEM images of fibres electrospun from solutions with different compositions . 65
vii
Figure 4.5: SEM image of fibres electrospun from 50 wt% SKL solution without PEO. a)
Purified SKL fraction F4. b) SKL without purification. (Scale bar = 20 m) ........................ 66
Figure 4.6: Mean fibre diameter vs. relaxation time () for F4/PEO solutions ..................... 67
Figure 5.1: Photograph of electrospun softwood Kraft lignin nonwoven fabric after
electrospinning ........................................................................................................................ 71
Figure 5.2: SEM images of electrospun fibres obtained from solutions containing (a) 32
wt% F1-3, 0.2 wt% PEO and (b) 32 wt% F4, 0.2 wt% PEO. Scale bar = 50 m. .................... 72
Figure 5.3: Photograph of electrospun lignin fabric after thermostabilization at 250oC in air.
Unfused fibres are shown........................................................................................................ 74
Figure 5.4: SEM images of electrospun fibres containing different ratios of F4/F1-3 after
thermostabilization at 5 oC/min to 250oC, in air. (a) F4/F1-3 = 100/0, (b) 70/30, (c) 60/40, d)
50/50. ...................................................................................................................................... 75
Figure 5.5: Thermostabilized (5oC/min) 50/50 F4/F1-3 film placed on moist paper (a-d), then
moved to dry paper (e-h)......................................................................................................... 77
Figure 5.6: (a) 3-dimensional AFM height image and (b) corresponding adhesion force map
of a 50/50 F4/F1-3 moisture-sensitive film heated at 5 oC/min. The size of the imaged area was
30 x 30 m .............................................................................................................................. 80
Figure 5.7: Adhesion force sections from AFM images on 50/50 F4 blend materials (top) and
a film containing only F1-3. ..................................................................................................... 81
Figure 5.8: Adhesion force superimposed on a height image of a moisture responsive film
with F4/F1-3 ratio of 50/50. ...................................................................................................... 82
Figure 5.9: Thermorheological responses of lignin F1-3 and F4 fractions. Average (n = 3) first
heat storage (G’) modulus (Top), and tan δ (Bottom) are presented. ..................................... 84
Figure 5.10: Dynamic rheology of electrospun fabrics of F1-3 and F4 blend (50/50). Effects of
heating rates on storage modulus (Top) and tan  (Bottom) are presented. ........................... 86
Figure 5.11: 1H-NMR of acetylated F4 (top) and F1-3 (bottom) from 2.6-1.7 ppm, showing
the peaks corresponding to acetylated phenolic (2.5-2.2 ppm) and aliphatic (2.2-2.0 ppm)
hydroxyl groups ...................................................................................................................... 88
Figure 5.12: 13C-NMR of acetylated F4 (top) and F1-3 (bottom) in the region 172-167 ppm,
corresponding to carbonyl carbons of acetyl groups .............................................................. 89
viii
Figure 5.13: Oxygenated aliphatic region (1H: 2.7-6.5 ppm, 13C: 48-95 ppm) of HSQC
spectra of F4 (top) and F1-3 (bottom). Unassigned peaks are traced in black. ......................... 91
Figure 5.14: 13C-NMR of F4 (top) and F1-3 (bottom) (a) etherified C-4 in guaiacyl units (b)
-O-4,13C ’, -5, 13CcC-O-4 C’; d) C-O-4, ; (e) methoxyl.
................................................................................................................................................. 92
Figure 5.15: Light scattering data showing elution curves obtained from GPC-MALLS for
acetylated fractions. Black: F1-3, 3 mg.mL-1, red: F4, 1 mg.mL-1, green: F4, 2 mg.mL-1, blue: F4
3 mg.mL-1 ................................................................................................................................ 95
Figure 6.1: (a) Non-bonded F4 fibres (NBF-250) and (b) Bonded 70/30 (w/w) F4/F1-3 fibres
(BF-250) after thermostabilization at 250oC. Scale bar = 20 m. .......................................... 99
Figure 6.2: (a) Non-bonded lignin-based carbon fibres (NBF-1000) and (b) Bonded (w/w)
F4/F1-3 carbon fibres (BF-1000) after carbonization at 1000oC. Scale bar = 5 m. .............. 100
Figure 6.3: Photographs demonstrating the differences in flexibility between (a) NBF and
(b) BF after carbonization (Tc = 600oC). Slight bending resulted in breaking of the NBF
material (a) while BF materials were relatively flexible (b). ................................................ 103
Figure 6.4: Typical Raman spectrum of carbonized lignin-based fibres in the wavenumber
region 900-1800 cm-1. The D-band (~1310 cm-1) is fitted with a Lorentzian line shape and the
G-band (1580 cm-1) is fitted with a Breit-Wigner-Fano (BWF) lineshape, both shown in
black, and the cumulative spectrum based on fitting is shown in red. .................................. 107
Figure 6.5: ID/IG from Raman spectra as a function of carbonization temperature for PAN,
NBF, and BF. Error bars represent plus/minus one standard deviation. PAN = ♦, NBF = ■,
BF = ▲. ................................................................................................................................ 108
Figure 6.6: Full-width at half-maximum (FWHM, cm-1) of the (a) D-band and (b) G-band as
a function of carbonization temperature from Raman spectra of PAN, NBF, and BF. PAN =
♦, NBF = ■, BF = ▲. Error bars represent plus/minus one standard deviation. .................. 108
Figure 6.7: Positions (cm-1) of (a) D-band and (b) G-band as a function of carbonization
temperature from Raman spectra of PAN, NBF, and BF. PAN = ♦, NBF = ■, BF = ▲. Error
bars represent plus/minus one standard deviation. ............................................................... 109
Figure A1: Raw data for plot of specific viscosity vs. concentration (Figure 3.5)………...151
Figure A2: Fitting to obtain slopes of specific viscosity vs. concentration (Figure3.5)...…152
Figure A3: Raw data for stress sweep (Figure 4.1a)……………………...………………..153
ix
Figure A4: Raw data for frequency sweep (Figure 4.2b)……………………......…….…..154
Figure A5: Example of CaBER data to obtain relaxation times (Chapter 4)…………..….155
Figure A6: Aliphatic region from HSQC of F4SKL in DMSO-d6………..……………….156
Figure A7: Aromatic region from HSQC of F4SKL in DMSO-d6………………...….…...157
Figure A8: Aliphatic region from HSQC of F1-3SKL in DMSO-d6…………………...…..158
Figure A9: Aromatic region from HSQC of F1-3SKL in DMSO-d6………………………..159
Figure A10: Wide angle X-ray diffraction patterns of BF (top) and NBF (bottom) carbonized
at different temperatures………………………………………………………………….…160
x
List of Symbols and Abbreviations
CF = carbon fibre
CNF = carbon nanofibre
AC = activated carbon
ACF = activated carbon fibre
SRM = stimuli-responsive material
SMM = shape memory material
SKL = softwood Kraft lignin
HKL = hardwood Kraft lignin
SOL = softwood organosolv lignin
HOL = hardwood organosolv lignin
PL = pyrolytic lignin
SL = sulfonated Kraft lignin
LS = lignosulfonate
DMF = N,N’-dimethylformamide
PEO = poly(ethylene oxide)
PAN = poly(acrylonitrile)
PVA = poly(vinyl alcohol)
NCC = nanocrystalline cellulose
SEM = scanning electron micrograph
F4 = softwood Kraft lignin fraction 4
F1-3 = softwood Kraft lignin fraction 1-3
 = average shear viscosity obtained from steady shear rheometry
s = viscosity of solvent determined by capillary viscometer at 25oC
sp = specific viscosity = ( – s)/s
G’ = storage modulus
G” = loss modulus
 = angular frequency in radians
|*()| = magnitude of the complex viscosity = √(G’2+G”2)
CaBER = capillary breakup extensional rheometer
xi
Dmid(t) = midpoint diameter of thinning fluid filament measured by CaBER
D1 = initial midpoint filament diameter after initial step strain during CaBER
λ = characteristic time scale of tensile stress growth in uniaxial elongational flow (relaxation
time)
G = elastic modulus
 = surface tension
Tc = maximum carbonization temperature
NBF = non-bonded fibres
NBF-250 = non-bonded fibres thermostabilized at 250oC
NBF-600 = non-bonded fibres carbonized at 600oC
NBF-800 = non-bonded fibres carbonized at 800oC
NBF-1000 = non-bonded fibres carbonized at 1000oC
BF = bonded fibres
BF-250 = bonded fibres thermostabilized at 250oC
BF-600 = bonded fibres carbonized at 600oC
BF-800 = bonded fibres carbonized at 800oC
BF-1000 = bonded fibres carbonized at 1000oC
 = electrical conductivity
SBET = Brunaer-Emmett-Teller specific surface area
FWHM = full-width at half-maximum
ID/IG = Ratio of the intensities of the D- and G- band measured with Raman spectroscopy
xii
Acknowledgements
I would like to first acknowledge and thank my research supervisor John Kadla for
bringing me into the Advanced Biomaterials Chemistry lab and supporting me for the last
five years, and for his patience in training me. His example has strongly shaped my approach
to science and will continue to do so throughout my career. I would also like to thank my
supervisory committee members, Professors Frank Ko, Savvas Hatzikiriakos, and Rodger
Beatson. My eyes are open to an endless world of possibilities to explore in the field of
polymer science because of them and I am eternally grateful for the guidance that I have
received during my time at UBC. I also must specifically thank several people from the
Biomaterials Chemistry lab who have helped me along the way. Dr. Yong-Sik Kim took me
under his wing near the beginning of my Ph.D program and prepared me for the long road
ahead. I am forever grateful that he took the time to train me in the lab and I am not sure if I
would have made it through the first difficult months without him. I’d also like to thank Reza
Korehei and Ana Filipa Xavier for their consistently positive, contagious energy which rarely
diminished over the last five years we have worked together. Their genuinely good nature
was especially comforting at those times over the last five years when things just did not go
according to plan, which happened often to say the least. I also wish to express my thanks
and best wishes to everyone else from Biomaterials Chemistry, AMPEL, chemical
engineering, and other labs and departments who helped me along the way.
I’d also like to thank my musical friends for providing much needed distraction from
the toils of Ph.D research. In no particular order and hopefully not forgetting any, they are
Peter Arcese, Athena McKown, Sierra Curtis-McClane, Jamie Leathem, Laurie Marczak,
Trevor Lantz, Joe Bennett, Chris Bater, Pete Cramer, Luc Desmarais, Rod Docking,
Christian Beaudrie, Isla Myers-Smith, Sarah Gergel, Paolo Segre, and the Steves. Thank you
all so much. I really enjoyed our time spent together and have each of you to thank for a
decent portion of my sanity.
Finally, I’d like to acknowledge generous support from North Carolina State
University/United States Department of Agriculture and the Natural Science and Engineering
Research Council of Canada. Research would not happen without the support of parties
dedicated to the advancement of science.
xiii
Dedication
I do not believe I could have completed this research without the love and support of my
family. This dissertation is dedicated to my parents Cynthia and Scott, my sister Emily, and
my brothers Bill and Rick. You were always there for me and it has made all the difference.
xiv
Chapter 1. Introduction
1.1
Motivation for development of advanced lignin-based nanofibre materials:
There is currently increasing interest in replacing non-renewable petroleum-based
fuels, chemicals, and materials with products derived from renewable resources.1–6 In theory,
biorefineries of the future will be able to meet the energy and material needs of society by
producing biofuels and other products from biomass such as wood and grass. Because lignin
is an integral component of biomass, biorefineries will inevitably produce it as a coproduct.1,5,6 However, lignin is currently under-utilized in spite of the fact that it is known
that utilization of lignin in value-added applications could benefit the overall sustainability
of biorefinery processes.6,7 Lignin is also currently generated in large quantities by industrial
pulping processes (predominantly Kraft pulping)8,9 which are difficult to measure precisely,
but were estimated in 1990 at 138.5 million kg/year worldwide.10,11 Gellerstedt and
coworkers9 have more recently estimated that assuming 10% of lignin from European Kraft
pulp mills were isolated, 1.5 million tons of Kraft lignin could be made available for valueadded products. However, as of 1998, only about 1% of all lignin generated in paper
production worldwide was isolated and sold,12 mostly as lignosulfonates. Most of the lignin
generated by current industrial processes is Kraft lignin from black liquor, which plays an
important role as a fuel in the energy balance of modern Kraft pulp mills. However, the
capacity of the recovery boiler is a bottleneck in the production of pulp.8,9 Removing some
lignin from black liquor prior to burning it in the recovery boiler could potentially increase
pulp production while also producing lignin for potential value-added applications.
Current applications of industrial (or “technical”) lignins include relatively low value
dispersants, emulsion stabilizers, surfactants, and binders.11,12 However, other higher value
opportunities for lignin utilization as a feedstock for chemicals and materials exist.7,13 A
large body of research has been conducted and considerable research activity is currently
underway in the area of chemical modification of lignin for use in polymer blends,
thermoplastics, polyurethanes, stimuli-responsive materials, hydrogels, and various resin
systems.4,14–27 Despite continued research on lignin utilization, commercialization of lignin1
based materials has proven difficult.11–13 The relatively limited utilization of industrial
lignins for many applications is related to its highly complex, heterogeneous chemical
structure and variability among different types of lignin which makes processing of lignin
and control over its physical properties very challenging. However, several decades of
research has allowed improvements in properties of lignin-based materials to be achieved by
clarifying the relationships between the complex structure and macromolecular properties of
different lignins, processability, and material properties. Advances in science and
technology coupled with a growing desire to use renewable resources in place of fossil
reserves present exciting new opportunities for the development of novel advanced ligninbased materials.
Electrostatic spinning, commonly referred to as electrospinning,28–30 is hereby
proposed as a promising strategy to novel materials from technical lignins. Electrospinning
is capable of producing continuous fibres with very small diameters in the range of
nanometers to microns from an enormous variety of materials.28–30 Electrospinning is
attractive due to its ability to produce continuous fibres with diameters from 10 nm to
several micrometers (1 nm = 10-9 m) using a wide variety of materials. Some examples
include PAN,31,32 cellulose acetate,33 keratin,34,35 alginate,36 chitosan,37 bombyx mori silk38,39
poly(methyl methacrylate) (PMMA),40 poly(L-lactide) (PLLA),41 as well as inorganic
materials.42 Due to its enormous versatility for processing a wide range of polymers,
electrospinning is a promising method to produce a variety of nanofibre-based materials,
including nanocomposites,43–46 fibrous catalyst substrates,47,48 drug delivery devices,49 tissue
engineering scaffolds,50,51, and nanowires.52,53 The electrospinning technique offers unique
advantages and presents exciting opportunities for development of lignin-based materials.
However, electrospinning of lignin presents a few challenges compared to synthetic
polymers which are related to the complex, branched, heterogeneous, and widely varying
lignin structure and physical properties. Chapters 3 and 4 will be aimed at characterizing the
behaviour of lignin during electrospinning and will place an emphasis on understanding the
relationship between the rheology of the spinning solution and the formation of electrospun
fibres. From there the possibility of using electrospun lignin fibres as precursors for ligninbased materials with interesting properties will be explored.
2
One class of materials that is currently receiving widespread attention for a variety of
applications is polymeric stimuli-responsive materials (SRMs). SRMs are often referred to as
“smart” materials due to the capability of changing their properties in response to external
stimuli (e.g. chemical, light, temperature, pH, external electric and magnetic fields), based on
the dynamic rearrangement of supramolecular polymer networks. The numerous exciting
properties and applications of SRMs have been reviewed extensively.54–60 A highly
interesting sub-class of SRMs are materials capable of fixing temporarily “programmed”
shapes and returning to their original shape in response to external stimuli such as
temperature,61 light,62 electricity,63 magnetism,64 or moisture.65–76 These SRMs are
commonly referred to as shape memory materials (SMMs). Polymer SMMs have received
increasing attention in recent years due to exciting possibilities in potential biomedical
applications77 as well as in sensors78 and actuators.79 Polymers intrinsically show shape
memory on the basis of entropic elasticity, and hence a wide variety of polymer SMMs can
be designed.80–82 The shape memory effect is due to the combined action of two or more
distinct phases or segments, where one phase acts as a “switch” capable of undergoing a
transition or change in conformation or mobility in response to a stimulus, and the other
segment remains relatively immobile and allows the material to “remember” its permanent
shape. The presence of a physically and/or chemically cross-linked network structure is
another important characteristic of SMMs.80–82 Numerous SMMs based on synthetic
polymers including polyurethanes65,69–71,73–75, polyvinyl alcohol,66,67 poly(e-caprolactone),83
and others80–82 as well as polymer blends84–87 and composite systems64,72,76,88 have been
reported. The advantage of using synthetic materials in design of SMMs is the versatility in
generating materials with well-defined switching characteristics89 and the ability to memorize
multiple shapes.88,90–92 Use of bio-based materials72,76,93–95 has also been reported in the
preparation of SMMs, but has received considerably less attention. The inspiration for design
of SRMs/SMMs has grown in relation with the observation of stimuli-responsiveness in
biological systems.55 Notably, while SRMs and SMMs have emerged as an important class of
novel advanced materials, the preparation of SRMs based on lignin has not received much
attention.23 Electrospinning has also been shown to be an effective method of preparing a
variety of interesting SRMs. The high surface area to volume ratio of electrospun fibres has
been reported to be advantageous for the preparation of SRMs with high sensitivity to
3
external stimuli.96 A portion of the research described in chapter 5 will therefore be devoted
to investigating the possibility of fabricating SRMs and SMMs through electrospinning of
lignin.
Carbon materials such as lignin-based carbon fibre (CF) and porous carbons such as
activated carbon (AC) and activated CF (ACF) are another particularly promising potential
application for lignin.9,97–108 The combination of high strength and stiffness coupled with
low density make high-performance CF an ideal reinforcement for high-strength, lightweight composite materials used in aerospace, automotive, marine, and sporting goods
applications.98,99,109,110 Carbon materials including CF and AC also have numerous
additional applications, including adsorbents,111 catalyst supports,112,113 biomedical
materials,114 electromagnetic shielding,115 and other electrical applications115 such as
electrochemical double-layer capacitors for energy storage or capacitive deionization,116–121
Li+-ion batteries,122–124 fuel cells,125,126 and dye-sensitized solar cells.127 Carbon nanofibres
prepared through electrospinning are being intensively investigated for potential application
in the areas listed above.128 Unfortunately, lignin is relatively difficult to process into high
strength CF, with continued research resulting in mechanical properties unsuitable for
structural composites.98–100,110 The tensile strength and modulus of lignin-based CF is still
relatively low compared to those of CF derived from synthetic poly(acrylonitrile) (PAN)based copolymers.110 On the other hand, superior mechanical strength is not required for all
applications, the relatively high cost of PAN is a limiting factor in its widespread
application,99 and there is a need to replace non-renewable petroleum-based materials such
as PAN with renewable precursors. Reduction of the fibre diameter below that of
conventional micron-sized fibres through electrospinning could improve the material
properties and expand the applications of lignin-based CFs. It is known that the mechanical
properties of carbon fibre increase with decreasing diameter.98,100,129,130 Decreasing the fibre
diameter also increases the available surface area for interaction with matrices thus
producing a composite with better shear strength.31 Furthermore, the increased specific
surface area provided by a reduction in the fibre diameter can benefit the numerous nonstructural applications of CF mentioned above. While lignin-based CF properties must be
improved further to compete with PAN and pitch-based CF in terms of mechanical
properties, combining an understanding of lignin properties with novel processing strategies
4
such as electrospinning may allow further improvement of properties and expansion of the
potential applications of lignin-based CFs to the other areas listed above. CF materials
obtained from lignin are therefore a second class of materials that will be investigated in
chapter 6.
In order to effectively apply electrospinning in the fabrication of novel lignin-based
nanofibre materials, it is necessary to first understand lignin. As an introduction to the
proposed research, the structure and isolation of lignin will be discussed first in section 1.2
and 1.3, respectively. The electrospinning process will then be discussed in detail in section
1.4. SRMs and SMMs will be then discussed in section 1.5 and the prospect of using lignin in
SRM/SMM fabrication will be presented. Finally in section 1.6, a discussion of CF
processing, structure and properties and previous research on lignin-based CF will be
presented
1.2
Lignin biosynthesis and structure:
Lignins are one of the three main constituents of the cell walls and extracellular
space of arborescent gymnosperms and angiosperms. Lignins serve as structural elements in
plant stems, imparting rigidity and resistance to impact, compression, and bending. In
addition, they facilitate the transport of water, nutrients, and metabolites, and provide
resistance against degradation by microorganisms.131 Native lignins are polyaromatic
heteropolymers derived from copolymerization of mainly three p-hydroxycinnamyl alcohol
monomers, referred to as monolignols.132,133 These are p-coumaryl, coniferyl, and sinapyl
alcohols, which differ in the number of methoxyl groups at the 3 and 5 positions of the
aromatic ring. Other monomers also take part in lignification to a lesser degree.
Hydroxycinnamates are one example of monolignols other than hydroxycinnamyl
alcohols.134 This is an important distinction to accurately represent the plasticity of
lignification and the variety of possible structures, but for a general explanation of lignin
structure, consideration of the three primary monolignols is illustrative. The primary
monolignols give rise to the p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S) units in the
lignin polymer through a combinatorial radical coupling process. Softwood (gymnosperm)
lignins consist of mainly G units with low levels of S and H units, while hardwood
5
(angiosperm) lignins consist mainly of G and S units with low levels of H units, and grasses
have comparable amounts of S and G units, with higher amount of H units compared to
hardwoods and softwoods.132,133
Figure 1.1: Lignin monomers (a,b,c) and subunits (d,e,f) (a) p-coumaryl alcohol (b) coniferyl alcohol (c)
sinapyl alcohol (d) p-hydroxyphenyl (H) subunit (e) guaiacyl (G) subunit (f) syringyl (S) subunit
The biosynthesis of lignin precursors from carbohydrates follows the shikimic acid
pathway, an essential biosynthetic pathway with phenylalanine and tyrosine as products. The
first evidence for the importance of this pathway was provided by Brown and Neish135 using
14
C-labelled shikimic acid and phenylalanine. These investigators demonstrated that these
labeled compounds were efficient precursors for lignins in wheat and maple. Numerous
subsequent studies indicate that the shikimic acid pathway is responsible for monolignol
biosynthesis.136 Phenylalanine is converted to the p-hydroxycinnamyl alcohol lignin
precursors through an enzymatically mediated series of reactions, which has been studied
intensively by many researchers.132,133,137–143 It should be noted that different pathways to the
p-hydroxycinnamyl alcohols are possible due to the fact that multiple isoforms of different
enzymes exist and are differentially expressed depending on the stage of development and
6
environmental cues.132,144,145 Also, additional complexity arises from the fact that pathway
intermediates may affect synthesis or activity of enzymes involved in the pathway.132
During lignification, monolignols are coupled in an end-wise manner to form the
lignin polymer. In gymnosperms and some angiosperms, monolignols are transported to the
cell wall from the cambial tissue in the form of 4-O--D-glucosides, which are presumed to
be the storage and/or transport form.146 After transport, monolignols are oxidized to phenoxy
radicals by the action of laccases and/or peroxidases.132,133,147 Dehydrogenation of a
monolignol leads to the formation of a phenoxy radical. Due to the ability of electrondelocalized radicals to couple at various sites, different linkage structures are possible133
(Figure 1.2). The predominant linkages are the -O-4 (-aryl ethers), which account for
roughly half the linkages in lignin, but other linkages, such as the -5 (phenylcoumaran), 55’ (biphenyl), 4-O-5’ (biphenyl ether), -’ (pinoresinol), dibenzodioxocin (combination of
5-5’/-O-4/-O-4), and -1 linked structures are also present in the lignin polymer. It is
important to note that the type of monolignol affects the possible outcomes of radical
coupling reactions. For example, the aromatic C5 position is available for coupling in
coniferyl, but not sinapyl alcohol. This explains why syringyl (S) lignins have a higher
proportion of -O-4 units compared to guaiacyl (G) lignins, because the aromatic C-5
position is unavailable for coupling. Following radical coupling, re-aromatization reactions
occur. In the formation of a -O-4 linkage, nucleophilic addition of H2O to either side of the
planar quinone methide intermediate results in two distinct stereoisomers, threo- (syn-,
RR/SS), and erythro-(anti-, RS/SR), which are present in kinetically controlled amounts.133
7
Figure 1.2: Characteristic linkages in lignin (a) -aryl ether (-O-4) (b) phenylcoumaran (-5) (c)
resinol (-) (d) biphenyl (5-5’) (e) biphenyl ether (4-O-5) (f) dibenzodioxocin (-O-4/-O-4/5-5’)
Much of our understanding of the mechanism of lignification comes from early
experiments on synthetic lignin-related polymers by Freudenberg.148,149 These types of
synthetic lignins are commonly referred to as dehydrogenation polymers (DHPs). The
results of DHP experiments support the hypothesis that the polymer grows by combination
of a radical at the phenolic end of a growing oligolignol with another radical of a
monolignol, most often at the monolignol  position. For example, the structure of DHPs
prepared by the dehydrogenative polymerization of coniferyl alcohol in the presence of
horseradish peroxidase (HRP) and H2O2 depended on whether the monomers were
introduced all at once (“Zulaufverfahren”) or gradually over a longer period of time
(“Zutropfverfahren”).136,148 In the first case, dehydrodimerization reactions between
monolignols are favored. In dehydrodimerization reactions the products always consist of
coupling of at least one of the monolignols at its -position. An enlightening observation
8
from DHP experiments is that the frequency of -O-4 units is lower when monomers are
added in bulk all at once compared to gradual addition. Gradual addition shifts the favored
outcome from dehydrodimerization to coupling of an oligolignol with monolignol radical
and results in a greater -O-4 content. Since the -O-4 linkage is the most abundant linkage
in lignin, these results suggest that the outcome of coupling reactions may be regulated in
planta by control of the monomer supply rate. Model experiments suggest that the outcome
of coupling reactions also depends on the rate of radical generation, the presence of
polysaccharides, and the presence of the growing lignin polymer chain.150–153 Since several
coupling outcomes with different probablities of occurring are possible between
monolignols and the growing polymer based on a combination of several factors, lignin
polymerization can be viewed as a combinatorial process resulting in a non-repeating
sequence of monomer units.132,133 The result of this combinatorial process is a polymer chain
with an enormous variety of possible chain sequences and conformations.
Through the mechanism described above, plant cells are able to deposit lignins with
different structure and orientation in different areas of the cell wall (P1, S1, S2) and middle
lamella (ML). Lignification is one of the final stages of xylem cell differentiation, with
lignin deposition occurring after carbohydrate deposition. The process begins at the cell
corners in the primary wall, spreading to the middle ML and S1 region of the secondary
wall, and finally throughout the secondary wall around the lumen. When lignification is
complete, the highest lignin concentration is in the ML region, but since the ML accounts
for a relatively small volume fraction of wood, the highest total amount of lignin is in the
secondary wall.154 A notable area of ongoing research is the precise nature of the initiation
sites for lignification, which may be related to the regulation and control of lignin structure.
Davin, Lewis and coworkers have suggested that arrays of dirigent proteins play a similar
role as that observed in lignan biosynthesis, by controlling the outcome of phenoxy radical
reactions in a regio- and stereo-specific manner155–157 and paving the way for a template
polymerization where existing lignin chains act as templates.153,158,159 It has been suggested
that a template polymerization mechanism would lead to a regularly repeating sequence of
monomer units, but this hypothesis remains to be fully tested experimentally.158,159
Nevertheless, the presence of initiating sites for lignification is clear. Keller and coworkers
have suggested a close functional relationship between glycine-rich protein and lignin
9
deposition during cell wall biogenesis in protoxylem cells.160 Ralph and coworkers have
demonstrated that ferulate polysaccharide esters actively participate with lignin monomers
in oxidative coupling pathways to generate lignin-ferulate-polysaccharide complexes during
cell-wall development in ryegrass.161 There is reason to speculate that such complexes might
play a role in lignin-polysaccharide crosslinking, which could also be involved in initiation
and/or control of lignification.134 In any case, the bulk of the experimental evidence points to
control of lignin deposition by spatial and temporal regulation of gene expression and
monomer supply,132,133 where lignification proceeds by a combinatorial process with
different monolignol radical coupling outcomes possible but with different corresponding
probabilities. While alternative hypotheses have been presented,158,159 the current prevailing
view of lignin structure is that of a branched, heterogenous polymer with an irregular
sequence of interunit linkages.
It is, however, quite interesting that while the complex process of lignin
polymerization and deposition appears to produce a non-repeating sequence of monomer
units which is sometimes described as random, the overall structure and orientation of lignin
in different areas of the cell wall is clearly not random. Donaldson studied lignification in
radiata pine by transmission electron microscopy (TEM) and observed that growing lignin
particles in the middle lamella and primary wall region form roughly spherical particles,
suggesting an isotropic character to the ML.162 On the other hand, lignin in the secondary
wall was observed to form elongated structures which followed the orientation of cellulose
microfibrils, showing that the architecture of the surrounding carbohydrate matrix exerts
constraints on lignin deposition which lead to an anisotropic orientation.162 Terashima and
coworkers have recently proposed a model of the nanostructural assembly of the secondary
wall163 in which bundles of cellulose microfibrils are surrounded by a lignin-carbohydrate
complex which follow the orientation of the bundle. Atalla and Agarwal showed using a
Raman microprobe that the aromatic rings are preferentially oriented parallel to the plane of
the cell wall.164 More recent studies using polarized infrared spectroscopy165,166 have shown
that lignin has a preferred orientation in the secondary wall. Polarized infrared spectra
collected with simultaneous dynamic tension applied to oriented thermomechanical pulp
sheets also showed that the orientation of lignin is related to its mechanical function.166
Interestingly, out-of-phase spectra obtained with the infrared beam polarized 90o to the
10
stretching direction showed that lignin participates in viscous dissipation of energy in
response to mechanical stress.166 Studies on the lignin-carbohydrate complexes in wood and
chemical pulps have also revealed that distinctly different types of lignin are preferentially
bound with different hemicelluloses.167,168 Taken together the results of various experiments
paint a very different picture of lignin than that of a random, isotropic “glue” simply holding
fibres together in wood. In contrast, lignin has a range of architectures ranging from more
isotropic in the ML to an anisotropic structure capable of facilitating dynamic rearrangement
of the cell wall matrix in response to mechanical load. As the properties of lignin-based
materials are inextricably linked with the physical properties of lignin, increasing
understanding of the structure-property relationships of lignin in wood may open new
opportunities for design of advanced lignin-based materials.
1.3
Isolation of lignin by chemical pulping:
The ever-expanding understanding of the complex nature of lignin essentially began
as an effort to deconstruct lignified wood tissues for conversion of various biomass
components into useful products. Traditionally, most of the useful products are derived from
the carbohydrate components (i.e. pulp and paper). On the other hand, the partially degraded
lignins generated as co-products of chemical pulping are relatively under-utilized, and
therefore represent an opportunity to improve the economics of biomass utilization while
simultaneously taking better advantage of an abundant, renewable feedstock.9 However,
additional difficulty in lignin utilization results from the fact that the complex lignin structure
is further chemically altered during isolation in a process-dependent manner. A variety of
industrial and experimental processes involve delignification of biomass and produce slightly
different “technical” lignins as co-products. The chemical reactions and structures of lignins
isolated by different methods have been investigated extensively.169–176 Currently the most
commercially important processes are chemical pulping methods, especially the Kraft or
sulfate pulping process.8,9,169,173 Several other methods of biomass processing such as soda,169
acid sulfite,169,172 and organosolv pulping,174–176 and pyrolysis177,178 also have important
applications in pulp and paper and potential renewable fuel applications (cellulosic ethanol,
11
bio-oil production), and generate lignin-derived co-products with different degrees of
structural modification and degradation depending on the process and conditions.
Chemical pulping processes are aimed at the selective removal of lignin in a way that
minimizes the degradation of the carbohydrate components, which are then converted to
useful products. Thorough delignification typically requires harsh conditions, such as high
temperature and pressure and high alkalinity or acidity which change the lignin structure. In
chemical pulping, lignin is dissolved by both incorporation of hydrophilic groups and
cleavage into smaller fragments. Several of the common pulping processes can be classified
as either alkaline, acidic, or neutral, with the majority of industrial chemical pulping being
alkaline. Alkaline processes include soda (where the main pulping chemical is NaOH), sodaanthraquinone (soda-AQ), and the predominant Kraft (sulfate) processes. Soda pulping is the
simplest alkaline process, but has a relatively low selectivity for lignin vs. carbohydrates.179
Delignification in the alkaline processes is carried out at temperatures ranging from 140170oC in an aqueous solution of NaOH with other additives included to favor lignin
degradation, reduce lignin condensation, improve lignin solubility, and/or reduce
carbohydrate degradation. Table 1.1 summarizes the reactive species in different pulping
processes in order of importance for each process.169
Table 1.1: Pulping processes and reactive species
Pulping Process
Soda
Kraft
Polysulfide
Alkaline sulfite
Neutral sulfite
Bisulfite
Acid sulfite
Reactive Species
OHOH , HS /S2 , S2O32-, Sn2-, CH3SOH-, HS-/S2-, Sn2-, S2O32-, CH3SOH-, SO32-, HSO3-,
OH-, HSO3-, SO32H3O+, H2SO3, HSO3H3O+, SO2, SO2.H2O, H2SO3
The reactions of lignin occurring in alkaline pulping processes include cleavage of
phenolic - and - ethers, cleavage of non-phenolic -ethers, demethylation, sulfonation, and
condensation. Quinonemethide intermediates are formed when phenolic substructures with
-carbinol or a corresponding -arylether group are heated with aqueous NaOH.169 The
12
reactivity of quinonemethides is due to their tendency to form more stable aromatic
structures by reaction with nucleophiles. In the presence of different additives, corresponding
nucleophiles react preferentially with quinonemethides, increasing delignification selectivity.
Different additives have been ranked in terms of their reactivity towards quinonemethides. In
order of decreasing reactivity the ranking is: Anthrahydroquinone (AHQ) > SO32- > HSO3- >
HS- > OH-.169
In acidic pulping processes, delignification is often carried out in aqueous solutions of
sulfur dioxide (SO2) in the presence of bases such as magnesium, sodium, potassium or
ammonium hydroxide. Pulping temperatures are in the range of 130-170oC.169 Sulfite
processes are classified based on the pH and the relative percentage of SO2 combined with
base as sulfite. Major chemical reactions during sulfite pulping are sulfonation, hydrolysis,
and condensation. Other minor reactions occur including oxidation, reduction,
rearrangement, dehydration, thiosulfation, and sulfidation.172 In acid sulfite pulping, -OH
and -ether groups are readily eliminated to form electron-deficient centers, carbonium ions,
which react with hydrated SO2 or HSO3- to produce sulfonate groups, or resulting in lignin
fragmentation by cleavage of -aryl ether bonds.169 The incorporation of sulfonate groups
increases the lignin solubility in the aqueous pulping liquor, facilitating delignification.
Alternatively, carbonium ion centers undergo condensation reactions with electron rich
aromatic carbons. Condensation reactions increase in frequency with decreasing pH.
Condensation is considered undesirable in sulfite pulping (as well as in alkaline pulping)
because it results in an increase in lignin molecular weight and decreased lignin solubility.
The lignin produced from acid sulfite pulping is commonly referred to as lignosulfonate.
Lignosulfonates differ from kraft lignins in that they are higher molecular weight180–182 due
to the relative stability of the -O-4 ethers in acid sulfite pulping. Lignosulfonates, unlike
kraft lignins, are also readily soluble in water due to the incorporation of hydrophilic
sulfonate groups, which is why they find application as water-based dispersants.
In addition to traditional alkaline and acid pulping processes, delignification can be
carried out by pulping in the presence of organic solvents (organosolv pulping).174–176,183
Solvent pulping has some advantages over more conventional pulping processes, namely
lower capital costs, smaller scale for economically attractive operation, significantly lower
environmental impact, and the production of sulfur-free lignins which have interesting
13
potential as value-added co-products.183 While organosolv pulps are of somewhat lower
strength than Kraft pulps, post-treatments can improve organosolv pulp properties.184
Recently, organosolv pulping has received increased attention due to its potential application
in the production of biofuel such as cellulosic ethanol.185–188 Numerous organic solvents have
been proposed for organosolv delignification (commonly in conjunction with acid or other
catalysts) either as solvents or in combination with water. These include methanol, ethanol,
n-butanol, acetic acid, ethylene glycol, and ethylene glycol methyl ether.174 Effective lignin
solvents have been characterized as having a Hildebrandt solubility parameter of 10.5-12.5
and satisfactory hydrogen bonding capacity.189 The dissolution of lignin in organosolv
processes is due mainly to the cleavage of -aryl ethers and to a lesser extent, arylglycerol-aryl ethers. The mechanism of -aryl ether hydrolysis involves a benzyl carbonium ion
intermediate which can react either with water, forming a benzyl alcohol, or with an alcohol
solvent, generating a benzyl ether. Condensation reactions can also occur, and as in Kraft and
sulfite pulping, these reactions impede delignification.174,175 Organosolv pulping was also
found to preferentially attack the middle lamella lignin in the initial stages of pulping which
explains the fact that fibre separation occurs at less than 50% delignification in organosolv
pulping.184 The order of topochemical preference for lignin removal from the middle lamella
region was given as follows:190 organosolv > acid chlorite > neutral sulfite > acid sulfite >
Kraft.
In summary, different isolation processes produce lignins with both differences and
similarities. Upon thermo-chemical treatment of biomass, carbohydrates and lignin can be
separated to varying degrees depending on the species and the type and severity of the
isolation process. The lignin that remains for potential conversion to value-added products
retains some structural features of native lignin structure such as -O-4, -', and -5
linkages, and phenolic and aliphatic hydroxyl groups. Still, the lignin structure is
considerably altered during its isolation. Lignin isolated from biomass is therefore highly
heterogeneous due to both inherent heterogeneity and additional isolation process-dependent
chemical modification. Technical lignin in its isolated state is a mixture of oligomeric and
polymeric fragments, and often contains inorganic impurities and residual carbohydrates.
Additional factors affecting conversion of lignin into value-added products are the large
range of molecular weights from 103-105 g/mol or higher180–182,191 (and high polydispersity
14
index Mw/Mn) and variability in thermal properties22,98,192,193 of isolated lignin, which
complicates processing of lignin into materials by solution or thermal means. In many cases
it is important to purify lignin to make it suitable for conversion to value-added products.
Purification can involve both a process to separate ash and carbohydrate from lignin as well
as fractionation of lignin into two or more separate fractions with greater homogeneity in
terms of chemical structure and physical properties. Fractionation of lignin by membrane
filtration of pulping liquor prior to lignin isolation,8,191,194,195 selective precipitation from
liquor or other solution,196 or solvent extraction of isolated lignin197–203 has been suggested
as a way of decreasing the heterogeneity of lignin. The goal of fractionation is to isolate
fractions with specific properties better suited for specific applications. While the
advantages and disadvantages of different schemes for dividing lignin into useful fractions
can be debated, there can be no general consensus on the best method of fractionation until
acceptable performance has been demonstrated for lignin-based products for specific
applications. Solvent extraction of isolated lignin will be used in this work mainly due to
experimental convenience and because it is proven in published literature that solvent
extraction of commercially available lignin can effectively eliminate carbohydrate-rich
fractions as well as isolate lignin fractions with different chemical structure, molecular
weight distributions, and in particular, thermal properties.197,198,202 In this case the objective
is to investigate how using lignin fractions with different properties can translate into
desirable processing characteristics (solubility, thermal softening) and final product material
properties. In general, a prerequisite for selecting the optimal fractionation scheme is that a
value-added product should add enough value to offset the cost of lignin isolation and
fractionation, and ideally a single fractionation strategy could produce feedstocks for
multiple value-added products. Here we attempt to focus on the possibility of making novel
advanced materials with an established fractionation protocol and leave development of
more efficient or cost-effective fractionation schemes for future researchers. Focus will also
be placed on using fractionation to improve the processability of lignin with new techniques.
Versatile processing strategies that can be applied to lignins derived from different types of
wood and different isolation and fractionation processes could be valuable pathways to
obtaining lignin-based materials with interesting properties. Electrospinning will now be
discussed in section 1.4 as such a strategy.
15
1.4
Electrospinning:
Electrostatic spinning, more commonly known as electrospinning, refers to a method
of producing fibres using electrostatic forces to process materials into fibres from solutions
or melts. Electrospinning is closely related to electrospraying, the use of electrostatic forces
to produce droplets. The first patent for electrospinning was granted in 1934,204 although
relatively few applications emerged until much more recently. The theory was developed
further in works by Taylor,205–207 and studied experimentally by Baumgarten.208 Reneker and
coworkers209,210 sparked increased interest in the technique in the 1990’s. Research in this
area has grown substantially since that time. Several reviews28–30,49,211,212 provide summaries
of the theory and research activities in the electrospinning field.
The experimental electrospinning setup is depicted schematically in Figure 1.3. The
basic apparatus consists simply of a spinneret (needle and syringe) containing the spinning
solution, a conductive collector placed at a fixed distance from the spinneret, and a voltage
source which is used to apply a potential between the spinneret and collector.
16
Figure 1.3: Schematic illustration of electrospinning process
As the applied electric potential is increased, a droplet of solution at the tip of the
spinneret will charge to a high potential and the repulsion of charges accumulating on the
droplet surface will stretch the fluid into a conical shape commonly referred to as a Taylor
cone. Above a threshold voltage on the order of kilovolts, a charged fluid jet is emitted from
the droplet and stretched toward the collector. After initially following a straight path along
the electric field direction (stable jet region), the jet undergoes an extremely rapid bending
instability which causes it to follow a spiralling path as it travels toward the collector. During
the stretching and bending of the jet, the large increase in surface area of the jet results in
rapid solvent evaporation, causing solidification of the jet. The resulting nanofibre is
deposited onto the collector as a nonwoven fabric consisting of randomly oriented fibre
segments. An example of a nonwoven fabric consisting of PAN fibres with diameters of
about 300 nm is shown in the scanning electron micrograph image in Figure 1.4.
17
Figure 1.4: Scanning electron micrograph of electrospun PAN nanofibre nonwoven fabric.
Scale bar = 50 m
The chaotic path of the electrospinning jets and the morphology of the resulting
nanofibres depends on the complex interplay between the electric field, which tends to
stretch the jet axially, surface tension, which tends to produce droplets to minimize surface
area, and the rheology of the solution under the stretching flow imposed on it. The different
stages of electrospinning have been studied in detail from a modeling and theoretical
perspective in several works.213–225 The small diameter of electrospun fibres results from
enormous elongational strain rates imposed by the electrospinning process estimated to be in
the 100-1000 s-1 range.213,226 The high rate of stretching also produces a high degree of
molecular orientation.32,227 As a result of this orientation, electrospun fibres are typically
birefringent.213
During electro-hydrodynamic jetting, axisymmetric instabilities (especially the
surface tension-driven “Rayleigh” instability) can cause the jet to either break into droplets as
in electrospraying, or promotes the formation of fibres with beads-on-a-string or cylindrical
morphology. Whether or not the jet breaks into droplets or forms fibres depends on the
balance between surface tension, which tends to favor droplet formation and axial stress
generated by capillary, viscoelastic, and electrostatic forces. The fibre morphology depends
18
on fluid properties such as viscosity, relaxation time, conductivity, and surface tension,
operating parameters such as applied electric field, collection distance, and solution flow
rate, and environmental variables such as temperature and humidity. The effect of various
parameters on the electrospinning process has been investigated theoretically and
experimentally in several studies.40,228–236 Since the fluid properties are functions of chemical
structures and interactions between the species present in solution, and fluid properties
greatly influence electrospinning, understanding the correlations between fluid properties and
electrospinning behaviour offers insight into the relationships between polymer structure and
processability by electrospinning.
In general, fluids consisting of low molecular weight polymers have a higher
tendency to form droplets, while entangled solutions of higher molecular weight polymers
tend to form fibres. Experimental investigations on electrospinning of synthetic polymers
indicates that entanglements between polymer chains are related to electrospinning
behaviour.40,230,231 Entanglements occur above a threshold concentration, sometimes denoted
Ce,230 which marks the boundary between the semi-dilute unentangled, and semi-dilute
entangled regimes.237–239 McKee et al.230 observed that the onset of beaded fibre formation
corresponded with the transition from the semidilute-unentangled to semidilute-entangled
regimes, while uniform fibres were typically obtained at concentrations roughly 2-2.5 times
Ce for a series of linear and branched polyesters with different molecular weights and degrees
of branching. They also observed that Ce increased with decreasing molecular weight and
increased branching, and that higher normalized concentrations (C/Ce) were required to
electrospin the lowest molecular weight polymers. These results demonstrate that
entanglements promote formation of electrospun fibres, fibre formation is favored by higher
molecular weight, and that higher concentrations are required for fibre formation to occur
with branched, low molecular weight polymers. Experiments by Shenoy et al231 and Gupta et
al40 using different synthetic polymer/solvent systems have also confirmed the important role
of chain entanglements in electrospinning. However, it has also been demonstrated that other
intermolecular interactions can stabilize the electrospinning process similar to entanglement
couplings even in the case of low molecular weight species. McKee et al showed that
hydrogen bonding in a series of poly(alkyl methacrylate)s240 or association in the low
molecular weight phospholipid lecithin241 can act like entanglements which can facilitate the
19
electrospinning process. Furthermore, it has been shown that entanglements are a sufficient,
but not necessary condition for fibre formation by electrospinning. Yu et al229prepared a
series of solutions containing low molecular weight poly(ethylene glycol) (PEG) and small
amounts of high molecular weight PEO in order to study the effect of elasticity on fibre
formation and morphology. The PEO-PEG solutions were prepared so that viscosity, surface
tension, and conductivity could be held constant while different degrees of elasticity could be
achieved by varying the molecular weight of the PEO component. Elasticity was
characterized by the relaxation time and steady extensional viscosity, which were measured
using a capillary breakup extensional rheometer (CaBER). It was emphasized that these
PEO-PEG electrospinning solutions behaved as Newtonian fluids in shear based on steady
shear measurements, and were considered to be un-entangled based on these observations.
Interestingly, when these fluids were subjected to electrospinning, the un-entangled solutions
could form bead-free fibres. The onset of fibre formation was related to elasticity,
characterized by the fluid relaxation time. They also showed that the elongational viscosity
increased as a function of relaxation time even though the shear viscosity remained constant.
This result emphasizes the discrepancy between different classes of polymeric liquids in
terms of the relationship between shear and elongational rheology. Thompson and
coworkers232 have also suggested that fluid relaxation time is one of the most important of 13
governing parameters in electrospinning investigated based on a theoretical model. These
results indicate that to correlate electrospinning behaviour and fluid properties for complex
systems, the relationship between shear and elongational fluid properties must be taken into
account. The results also emphasize the fact that there is a complex relationship between
molecular structure and interactions between different fluid components, viscoelastic fluid
properties, processing parameters, and the morphology of electrospun fibres.
In the studies on electrospinning of entangled, synthetic polymer solutions mentioned
above, zero-shear viscosity 0, obtained by steady shear rheometry, was the parameter used
to determine information about the extent of entanglement in different systems. This is
logical in light of the well-known scaling relationships which can be used to deduce
information about polymer chain entanglements in polymer melts and solutions. It is also
clear that fibre diameter tends to correlate with 0, and typically the transition from spray to
beaded fibre to uniform fibre correlates with increasing 0. While important correlations can
20
be drawn from shear viscosity data, the rheological behavior of viscoelastic polymer
solutions in elongational flow can not be predicted from rheological measurements in shear,
particularly in the case of large, rapid deformations, in which case nonlinear viscoelasticity
must be considered.219,239 There is also a countless number of possible systems with multiple
components blended together in solution, broad molecular weight distributions, different
degrees of polymer branching, and different types and strengths of intermolecular
interactions. Extrapolation of correlations obtained for systems consisting of entangled,
linear, homopolymers to more complex systems is therefore problematic. However, given the
known fact that numerous parameters influence electrospinning it is simply not practical to
approach each electrospinning system as a process of trial and error in order to obtain the
smallest diameter fibres which are free of beads-on-string defects, and hence it is essential to
understand the rheology of the electrospinning solution in order to exert control over the fibre
diameter and morphology.
These facts must be taken into account when considering using lignin in the
formation of electrospun fibres, because it has been demonstrated that lignin is relatively
difficult to process by electrospinning compared to synthetic, linear polymers.
Electrospinning of Alcell lignin using a coaxial spinneret system has been reported.242,243 In
these investigations, a sheath layer of solvent was necessary to stabilize the electrospinning
process in order to generate uniform fibres. Below a certain sheath fluid flow rate, Alcell
lignin solutions were observed to electrospray. Other researchers have studied
electrospinning of sulfonated alkali lignin blended with nanocrystalline cellulose
(NCC)/poly(vinyl alcohol) (PVA)244 and PAN.245 The results of these experiments
demonstrate that lignin can be processed into nanofibres by electrospinning. However, there
are limitations to be addressed with respect to each of these studies. Each published study
only reported on electrospinning of a single type of lignin. Questions therefore remain on
whether the strategies will be effective for a variety of technical lignins, which vary widely in
structure and properties. Alcell lignin is the only type of lignin so far reported to be capable
of forming fibres without the addition of synthetic polymers in the spinning solution.242,243
However, no correlation of the rheological properties of Alcell lignin solutions and
electrospinning behaviour has been reported. The relationship between lignin in solution and
fluid rheology are therefore not well understood. Some shear viscosity measurements have
21
been reported in the study of lignin/NCC/PVA and lignin/PAN systems, but no in-depth
study on viscoelasticity of lignin-based solutions has been reported with any rheological
characterization other than steady shear. Furthermore, while the Alcell lignin nanofibres were
processed into carbon nanofibres and the surface properties were measured, very slow
stabilization heating rates were required to stabilize the fibres before thermostabilization, and
the mechanical and electrical properties of Alcell lignin-based CNFs were not reported.
Lignin/NCC/PVA nanofibres were not carbonized, and were not characterized in terms of
their mechanical properties, hence the effect of lignin content and NCC reinforcement was
not clear.244 Lignin/PAN nanofibres cured with electron beam irradiation were carbonized
and characterized in terms of their mechanical properties, but the only mechanical properties
reported were for fibres with lignin content of 50%, thus it is difficult to ascertain the effect
of lignin on the mechanical properties of the fibres. Also, even though it is known that PANbased CNFs have promising potential for use as carbon electrodes, the electrical conductivity
of lignin-PAN CNFs was not reported. Greater investigation into the fundamental
requirements for formation of electrospun fibres from lignin as well as investigations aimed
at preparing and characterizing electrospun lignin-based fibres should be undertaken to
evaluate the potential for obtaining interesting lignin-based materials through
electrospinning.
1.5
Stimuli-responsive materials from lignin:
Shape memory materials (SMMs) are a class of stimuli-responsive materials (SRMs)
which can undergo reversible changes in shape in response to external stimuli. The shape
memory effect has been described in both metal alloys and polymers. The one-way shape
memory effect was first observed in metal alloys by Chang and Read246 for a gold-cadmium
alloy, followed in 1963 by the observation of shape memory effect in nickel-titanium alloy
by Buehler et al,247 now well-known as nitinol. The mechanism of shape memory in NiTi
alloys is a transformation between martensitic and austenitic crystalline phases upon
heating.61 Polymer shape memory was observed even earlier, described in a US patent by
Vernon in 1941.248 The commercial relevance of shape memory polymers was realized in
the 1960’s with the introduction of “heat-shrinkable” cross-linked polyethylene,249 and
22
research activity expanded thereafter and increased with the introduction of shape memory
polyurethanes. Polyurethanes are the most widely studied class of polymer SMMs, first
discovered in the 1980s at Mitsubishi Heavy Industries.250
The most common SMMs are thermally activated, in that they have the remarkable
ability to remember a “programmed” shape which is determined by first heating the material
above a transition temperature, then deforming it into the desired programmed shape, and
subsequently cooling it back below the transition temperature without removing the imposed
strain. The ability to remember programmed shapes depends on the presence of at least two
distinct phases or segments which can fulfill specific roles in the shape memory property.
One is a relatively rigid, immobile component which maintains its mobility or conformation
upon application of a stimulus. The other phase acts like a molecular switch which changes
its mobility in response to an external stimulus.61 Using the example of a linear polyurethane
with a block copolymer structure and thermally induced shape memory effect, the immobile
“hard” segment would have a higher glass transition (Tg) or melting (Tm) temperature, while
the switch or “soft” segment would have a lower Tg/Tm. The chemical structure of the
different segments should also be incompatible enough so that the material displays a phase
separated morphology.61,70 Upon heating the material above the transition temperature of the
soft segment, this component would become rubbery/viscous, while the hard segment
remains elastic. Deforming the material then results in storage of elastic energy and
stretching polymer chains in the hard segment into an extended, entropically unfavorable
state, while the soft segment can effectively reorganize into an entropically favorable state to
accommodate the imposed deformation. If the imposed strain is maintained and the material
is subsequently cooled below the transition temperature of the soft segment, the soft
segment returns to an immobile glassy state, effectively freezing the hard segment in a state
of relatively low entropy due to confinement in its deformed or stretched conformation. The
elastic energy stored during deformation then remains stored in the hard segment after
cooling. Shape recovery can then be activated once the material is reheated above the soft
segment transition temperature and the imposed strain is removed, allowing the hard
segment to return to the coiled state (maximum entropy).
SMMs can also be designed to respond to a variety of stimuli, including light,62
electricity,63 magnetism,64 or moisture.65–76 The possibilities for the design of specialized
23
SRMs and SMMs are therefore very promising. In essence, stimuli-responsiveneness and
shape memory properties depend on the ability of a supramolecular polymer network to
rearrange itself on application of a stimulus and return to its original state once the stimulus
is removed.57,60 Recent advances in the field of SRMs/SMMs have taken advantage of the
variety of different physical/noncovalent and covalent cross-links that can be employed to
influence the stability of supramolecular networks.57 For example, “triple-shape” SMMs
capable of memorizing a permanent shape and two programmed shapes can be prepared by
introducing multiple segments with different transition temperatures, which is essentially an
extension of the mechanism of thermally activated “one-way” SMMs.88 Noncovalent
interactions such as self-complementary hydrogen bonding can also be exploited to prepare
advanced triple-shape SMMs.91 Hydrogen bonding has also been shown to be a key aspect
of the mechanism in many moisture-responsive SMMs. It was accidentally discovered that
shape recovery in thermally activated polyurethane SMMs could also be activated by
exposing the material to moisture.75 It was later determined that the underlying mechanism
was the disruption of the network of hydrogen bonds in the material by bound water, which
increased the mobility of the molecular chains of the switching phase and decreased the
Tg.69,70,74 Moisture-responsive composite SMMs have also been prepared using
nanocrystalline cellulose embedded in a polymer matrix.72,76 Strong hydrogen bonding
between cellulose crystals could be induced by drying or disrupted by introducing water,
resulting in a mechanically adaptive, moisture-responsive SMM.72,76
The use of cellulose in the preparation of SMMs via the reversible disruption and
formation of hydrogen bonded networks raises the interesting question of whether lignin
might also be used in preparing moisture responsive SMMs. The mechanical properties of
wood are governed by a variety of covalent cross-links between the major components,
cellulose, hemicellulose and lignin, as well as noncovalent interactions. Decades of research
have been devoted to understanding of cellulose and the network of hydrogen bonding
governing its properties.251 Wood cell walls are composites of stiff cellulose microfibrils
running along the fibre direction, embedded in a highly oriented amorphous matrix of lignin
and hemicelluloses.252,253 It has been reported that a never-dried wood cell walls have a
remarkable ability to undergo plastic deformation without serious damage to the material.
Keckes et al.254 have described this behavior as a “molecular Velcro” or “stick-slip”
24
behavior similar to moving dislocations in metals. Hill and coworkers255 used proton spin
diffusion NMR to show that the stick-slip mechanism is consistent with the presence of
moisture mediating the degree of hydrogen bonding between cellulose microfibrils and the
surrounding matrix (hemicellulose and lignin). Compared to our understanding of the role of
cellulose and hemicellulose,256 the role of lignin in governing this fascinating mechanical
behaviour is not well understood. However, it is known that the molecular mobility of lignin
is strongly increased by the presence of water.22,193 It is also known that different types of
noncovalent interactions govern the physical properties of lignin, including hydrogen
bonding257 and nonbonded orbital interactions between aromatic rings.258–260 Furthermore,
studies on the lignin-carbohydrate complexes in wood and chemical pulps have also
revealed that distinctly different types of lignin are preferentially bound with different
hemicelluloses.167,168 Based on studies on wood and pulp, it appears that dynamic
reorganization of lignin and/or lignin-carbohydrate complexes on a supramolecular level in
response to the environment might be of fundamental importance in nature. Reorganization
of the lignin-hemicellulose matrix may also play a role in fibre curling in pulp-making,
referred to as latency.261 Interestingly, mechanical pulps which are typically somewhat
higher in lignin content can be uncurled by heating, similar to a temperature sensitive SMM.
Reversible reorganization of supramolecular networks is critical in determining the
behaviour of SRMs and SMMs. If different types of lignin in wood have different types and
strengths of intermolecular interactions governing their respective mobility, and these
differences in native lignin are still reflected in the properties of isolated lignin, these
intrinsically different characteristics might be useful in the design of lignin-based phase
separated SRMs or SMMs. Furthermore, it has been shown electrospinning is a highly
useful technique in the preparation of SRMs, as the interaction of materials with external
stimuli is often benefited by the high surface area to volume ratio of electrospun fibres.96
Therefore, electrospinning of lignin is a potentially promising route to novel SRMs and/or
SMMs based on lignin. This hypothesis will be further addressed in chapter 5.
25
1.6
Production of carbon fibres from lignin:
As mentioned in the introduction, carbon materials such as carbon fibres (CFs) and
porous carbon materials such as activated carbon (AC) and activated CF are also promising
value-added products that can be prepared from lignin. CFs have a variety of applications in
reinforcement of composites for aerospace, transportation, sporting equipment, and marine
applications,98,99,109,110,262–264 and can also be used as adsorbents, catalyst substrates,
electrodes, and chemically resistant materials.111–127,265–272 CFs can be prepared from a
variety of precursor materials, most often PAN, petroleum pitch, or rayon109,110,273,274 but
also from other carbon rich precursors such as phenolic resin275,276 and lignin. The process
for making CF consists of fibre spinning, such as melt, wet, or dry spinning, followed by
thermal treatment under different conditions, generally low temperature oxidative
thermostabilization, which depends on the precursor, but is in the 200-400oC range.110
Stabilization cross-links and oxidizes the fibres and allows them to resist melting and
maintain the fibre shape during carbonization. Low temperature stabilization is followed by
carbonization at temperatures up to 1600oC under inert conditions, and in some cases
graphitization treatment at temperatures up to 3000oC, which increases the Young’s modulus
and electrical conductivity.109,110,263,277,278
The structure and properties of CFs depend on the precursor and spinning method,
heating temperatures and rates, as well as post-spinning steps to increase the orientation of
crystallites along the fibre axis109,110,279 or increase the surface area for applications such as
adsorbents and electrodes (activation). High modulus CFs are typically prepared from PAN
by wet or dry spinning or mesophase pitch by melt spinning and generally have an axially
oriented graphitic structure, with different transverse textures. These include onion-skin,
radial, flat-layer, and random textures observed in mesophase pitch-based CF and turbostratic
texture observed in PAN-based CFs.109,263 The excellent mechanical properties and low
density of PAN and mesophase pitch based CF make them useful for fibre-reinforced
composite materials.110 As examples of high modulus CF, Toray Industries M60J is a PANbased CF with Young’s modulus of 585 GPa,280 and several mesophase pitch based CFs are
reported263 to possess moduli over 800 GPa. High strength CFs are produced from
26
copolymers of acrylonitrile with low amounts of other comonomers such as itaconic acid,
which are included to initiate cyclization of nitrile groups and control the exothermicity of
PAN stabilization.109,277–279,281–285 The tensile strength of PAN-based CF increases up to
about 1600oC, but unlike the modulus, tensile strength of PAN-based CF decreases at higher
carbonization temperatures.110,263 The high strength of PAN-based CFs may be due to
turbostratic structure, which is less prone to defect-induced failure compared to more
graphitic structures characteristic of mesophase pitch.110,263 An example of a commercially
available CF with very high tensile strength is Toray Industries T1000G, which has a tensile
strength of 6370 MPa.286
Lignin has been under investigation for its potential as CF precursor since the
1960's.287 The first commercial lignin CF, Kayacarbon, was produced by Nippon Kayaku,
Co. Subsequent attempts have produced CF from lignin,288–291 but the mechanical properties
of lignin-based CFs produced to date are generally not comparable to CFs produced from
PAN or mesophase pitch.9,99,110 The preparation and properties of a variety of lignin-based
CFs have been reviewed by Gellerstedt, Sjoholm, and Brodin, and by Kubo and Kadla.9,97 In
addition to being relatively weak, the processability of lignin into fibres depends on the
source and type of pulping process used for isolation of lignin.9,97,98 In some cases it was
necessary to introduce pretreatment steps to generate a fusible precursor material from lignin
which was capable of forming fibres by thermal extrusion.288 Dave, Prasad, Marand and
Glasser292 emphasized that in order to produce CFs with good mechanical properties, the
precursor must have the ability to form a fluid, yet organized, liquid-crystalline state when
spun into fibres in order to generate a highly oriented graphitic structure upon thermal
treatment. This type of behaviour is characteristic of mesophase pitch274 and interestingly,
was demonstrated for some lignin model compounds.292 However, technical lignins typically
form fibres with structure closer to that formed by isotropic pitch, which is presumably
related to their relatively poor mechanical properties. Impurities such as inorganic salts and
carbohydrates are also an important issue for lignin-based CF production.9,97–99 It is known
that defects must be avoided in order to produce a high strength fibre,156 as they result in
discontinuities in the carbon lattice of CF.97–99,293 In spite of difficulties in generating ligninbased CF with mechanical properties suitable for structural applications, research in this area
is ongoing.9,99,110
27
Blending of lignin with synthetic polymers has been demonstrated as one promising
route for improving the processability of lignin and properties of lignin-based CF for real
applications. General grade CFs have been produced from technical lignins (hardwood Kraft,
Alcell) and blends of lignin with synthetic polymers including and poly(ethylene
terephthalate) (PET), polypropylene (PP), and poly(ethylene oxide) (PEO).98,100,102 The
polymers were first blended by thermal mixing, extruded into sticks, and subsequently
formed into fibres by a second thermal extrusion. The fibres were then subjected to a twostep thermal processing sequence, typical for CF.110 The thermal treatment sequence
consisted of oxidative thermostabilization, where fibres were heated at a slow rate of
0.5oC/min up to 250oC and held for one hour under air, and subsequent carbonization by
heating at 3oC/min to 1000oC under N2.98,100 By this method, hardwood Kraft lignin (HKL)
fibres were processed into CFs with tensile strength of 605 MPa and Young's modulus of 61
GPa, while blends of hardwood kraft lignin with 25% PET could produce fibres with strength
and modulus values of 703 MPa and 94 GPa, respectively.100 Blending with lignin with PP in
a lignin/PP ratio of 63/37 yielded a porous CF with fairly high specific surface area of ~500
m2g upon thermal degradation of the PP component.102 It is interesting to note that although
this surface area is lower than typical activated carbons, no activation process was used in
this study. On the other hand, activated carbons and activated CF have been prepared from
lignin.101,103–108 Given the numerous exciting applications for carbon materials with high
surface area mentioned above, the surface area of lignin-based CF is an important parameter
to consider for development of lignin-based CF for real applications.
Blending with PEO was observed to enhance the spinnability of lignin during thermal
extrusion.98 The processability of the blend system was reportedly enhanced by miscibility
between lignin and PEO, which was due to the formation of hydrogen bonds between the
hydroxyl groups of lignin and the ether groups in the PEO backbone.98,192,257,294–296 An
important point to note is that the degree to which blending with PEO could enhance thermal
spinning of lignin was dependent on the type of lignin. Compared to HKL and Alcell lignin
(also from hardwood), softwood Kraft lignin (SKL) was observed to require higher amounts
of PEO (more than 50%) in order to obtain a fibre at high take-up speeds, and continuous
spinning could not be achieved at PEO contents of 25% or less.294 Since the mechanical
properties decrease and thermostabilization is hindered by high PEO content,98,192 blending
28
lignin with these relatively high amounts of PEO is impractical for SKL processing. The
thermal spinnability of technical lignins is strongly related to their thermal softening, which
is in turn related to their chemical structure. Hardwood lignins were previously observed to
display relatively good thermal spinnability even in the absence of PEO,98 while the
spinnability of SKL was poor due to a much lower tendency to soften and flow upon heating.
The low thermal mobility of SKL is believed to be due to a more highly cross-linked and
condensed structure.98,290 On the other hand, a disadvantage of the high thermal mobility of
hardwood lignins is that hardwood lignin fibres require slow heating rates during
thermostabilization and carbonization to prevent them from fusing together, which greatly
increases the processing times.98,99 Strategies for fibre production are therefore needed which
can be applied to different types of lignin and which overcome the inherent variability in
thermal properties between softwood and hardwood lignins.
Furthermore, the smallest diameter achievable through thermal spinning is in the
micron range. This is significant because CFs exhibit considerable diameter dependence in
their mechanical properties attributed to increased molecular order along the fibre axis,
reduced amount and size of defects, and improved uniformity of heat
treatment.129,130,277,279,32 Accordingly, increasing tensile strength has been previously
observed to increase with decreasing diameter for lignin-based CF, although the diameter
achieved using hardwood kraft lignin was in the 25-50 micron range.98,100 However, smaller
filament diameters on the order of 10 microns have been reported without any increase in
mechanical properties.99 Reduction of the fibre diameter could potentially increase the
mechanical properties of lignin-based CF, but it is still not entirely clear to what extent.
Even in the absence of increased single fibre mechanical properties, decreasing the fibre
diameter could also improve the interaction of lignin-based CF with matrices used for CFreinforced composites.31 In addition, reducing the fibre diameter could expand the
applications of lignin-CFs by increasing specific surface area. Several previously mentioned
CF applications (catalyst substrates, electrodes, adsorbents) utilize high specific surface area
or “activated” CF. There has been a surge in interest in multifunctional carbon materials for
applications related to energy storage, water treatment, and other areas.111–127 A versatile
strategy for processing of a variety of technical lignins with a range of chemical structure
29
and properties into very small diameter fibres could therefore open up a host of new
opportunities for lignin-based CF utilization.
In light of the relative abundance of publications on the properties and exciting
applications of electrospun PAN-based CNFs, the relatively high cost of PAN and the fact
that it is not a renewable material, and the relatively few reports on electrospinning of lignin,
the production and characterization of lignin-based nanofibres and CNFs by electrospinning
is a very promising avenue for the fabrication of new lignin-based materials. The first
important step to realizing the potential of electrospinning in generating new ligninmaterials is to develop a versatile system which can be applied to technical lignins with a
variety of structure and properties. In addition, the rheological phenomena underlying the
balance between droplet formation and production of uniform fibres must be elucidated in
order to control the electrospinning of lignin. The key research questions to be answered in
these areas are: 1) Can a simple strategy be employed to reliably obtain electrospun fibres
from different types of technical lignin?, and 2) What rheological parameters govern fibre
formation and the fibre diameter and how can they be tuned? These first two points will be
addressed in chapters 3 and 4. Having established a robust system for the production of submicron diameter lignin fibres by electrospinning, the effect of lignin structure and properties
on thermostabilization, carbonization, and the material properties of the corresponding
thermostabilized and carbonized nanofibres must be studied. Chapter 5 will focus on
oxidative thermostabilization of electrospun Kraft lignin nonwoven fabrics and the effect of
lignin structure on thermal mobility and inter-fibre fusion. Chapter 5 will also present the
exciting finding that electrospun lignin fibres can be converted to novel SRMs with simple
heat treatment of fibres containing certain amounts of lignin fractions with different
structure and properties. Finally, chapter 6 will focus on the mechanical properties, electrical
conductivity, and surface area of carbonized Kraft lignin nanofibres, and discuss inter-fibre
bonding as a means to enhance material properties. A detailed characterization and
comparison of the structural changes occurring during carbonization of Kraft lignin and
PAN nanofibres using Raman spectroscopy will also be presented. Before discussing the
research findings in each of these areas, Chapter 2 will present a detailed explanation of the
experimental procedures used to conduct this research.
30
Chapter 2. Materials and experimental methods
2.1
Materials:
Softwood kraft lignin (Indulin-AT, SKL), hardwood kraft lignin (HKL), and
sulfonated kraft lignin (Kraftsperse 25M, SL) were obtained from MeadWestvaco (Glen
Allen, VA, USA). Kraft lignins were washed with dilute HCl (pH 2) to remove water-soluble
impurities and ash. The lignin was suspended at 100 g/L stirred for 30 minutes, and filtered.
This process was repeated a total of five times. SKL and HKL were freeze-dried before
solution preparation and electrospinning in the case of electrospinning of unfractionated
lignin (Chapter 3). When additional fractionation steps were included (Chapters 4, 5, and 6),
SKL was not freeze-dried, but was instead dried under ambient conditions followed by
drying at 105 oC overnight. Hardwood organosolv lignin, Alcell (HOL) was obtained from
Aldrich (Oakville, ON) and used as received. Softwood organosolv lignin (SOL) was
prepared by a previously reported protocol.187 Pyrolytic lignin (PL) was precipitated from
bio-oil (Dynamotive Energy Systems, Vancouver, BC) and subjected to an initial heat
treatment at 160oC for 30 minutes under reduced pressure to remove volatile contaminants as
previously described.177 Lignosulfonate (Starflo dye dispersant, LS) was obtained from
Georgia Pacific (Bellingham, WA).
Poly(ethylene oxide) (PEO) with an average molecular weight of 6 x 105 g/mol was
obtained from Union Carbide (Houston, TX) and used as received. PEO with a nominal
viscosity average molecular weights Mv of 1 x 106 and 5 x 106 g/mol were obtained from
Sigma-Aldrich, and used as received. Polyacrylonitrile (PAN, Mw = 150,000) was obtained
from Scientific Polymer Products, and used as received. N,N-dimethylformamide (DMF),
methanol, and methylene chloride were all ACS Reagent grade and obtained from Fisher
Scientific (Ottawa, ON), and used as received.
31
2.2
Fractionation of SKL by sequential solvent extraction:
Fractionation of commercially available softwood Kraft lignin (SKL, Indulin-AT,
Meadwestvaco, Glen Allen, VA, USA) was carried out by sequential extraction with organic
solvents, based on a modification of a published procedure.197,198 The SKL was first washed
with dilute HCl (pH 2) to remove water-soluble impurities and ash as described above. The
resultant acid-washed SKL was then extracted with methanol by vigorous stirring for 30 min
at 100 g/L. The mixture was then filtered and the insoluble fraction was air-dried overnight,
and ground with mortar and pestle; this process was repeated a total of three times. The
methanol-soluble fraction (herein denoted F1-3) was concentrated on a rotary evaporator at 20
o
C under reduced pressure; the temperature was then increased to 50 oC for 30 min to
completely remove the solvent. The isolated F1-3 was then ground with a mortar and pestle to
a fine powder and subsequently dried for another 2 hours on a Schlenk line. The remaining
insoluble material after methanol extraction was air-dried, ground, and further extracted with
a 70/30 (v/v) mixture of methanol/methylene chloride, again the insoluble material was airdried, ground, and re-extracted a total of three times. The soluble material from the 70/30
methanol/methylene chloride washing (herein denoted F4) was concentrated and dried in the
same manner as F1-3.
2.3
2.3.1
Lignin characterization:
Thermorheological analysis of F4 and F1-3:
Thermorheological analysis was performed under dynamic compressive-torsion
mode57 on a TA Instruments (Grimsby, ON) AR2000 stress controlled rheometer. Finely
ground lignin specimens were held between two parallel plates (25 mm dia.) with a normal
force of 3±0.1 N, while a small amplitude dynamic torsional shear was applied (strain: 0.1%;
frequency: 1 Hz). Specimens were heated at 3 °C/min from 25 to 230 °C. All experiments
were under dry N2 and liquid N2 was used for temperature control. The instrument was
programmed such that, prior to material softening below a critical modulus the instrument
controlled the normal force over the specimen and switched over to gap control once the
32
shear moduli dropped below a critical value, to eliminate specimen compression. Electrospun
fabrics were analyzed with similar normal force and gap control. Circular discs of 25 mm
diameter were punched out and 3 such discs were stacked during the testing. To compare the
effects of heating rates, heating scans (25 to 250 C) were performed at different heating
rates, 1, 2 and 3 °C/min.
2.3.2
Acetylation of lignin:
Lignin fractions F4 and F1-3 were acetylated for NMR and GPC-MALLS
characterization; 200 mg of lignin was dissolved in 8 mL of pyridine/acetic anhydride (1:1,
v/v) and stirred for 48 h at room temperature. The reaction solution was then added dropwise
to 300 mL ice-water with stirring. The precipitated lignin was collected by filtration through
a Nylon membrane (0.45 μm, 47 mm), washed with ice-water, air-dried, and further dried on
a Schlenk line.
2.3.3
Characterization of lignin by nuclear magnetic resonance (NMR) spectroscopy:
1
H and 13C NMR were measured using a Bruker Avance 300 MHz spectrometer
equipped with a BBO probe. For quantitative 1H NMR, precisely 5 mg of acetylated lignin
was dissolved in 0.5 mL of CDCl3. 1 mg of p-nitrobenzaldehyde was also accurately weighed
and added to the NMR tube as internal standard. Phenolic and aliphatic hydroxyl group
contents were determined from the integration of acetoxyl protons located in the region 2.51.8 ppm using the internal standard aldehyde proton peak integral set to 1. The NMR spectra
were recorded at 300 K, with a 90° pulse width and a 1.3 s acquisition time. A 7 s relaxation
delay (d1) was used to ensure complete relaxation of the aldehyde protons. A total of 128
scans were collected.
Quantitative 13C NMR spectroscopy was performed using 200 mg lignin samples in
0.5 mL of DMSO-d6. Relaxation was facilitated by the addition of chromium acetylacetonate
(final concentration = 10 mM).297 Conditions for quantitative 13C analysis included a 90°
pulse width with a 1.4 s acquisition time and a 1.7 s of relaxation delay (d1). A total of 20000
scans were collected.
33
2.3.4
Characterization of lignin molecular weight distribution by gel permeation
chromatography and multi-angle laser light scattering (GPC-MALLS):
The molecular weight distribution of the acetylated lignins SKL, HKL, SOL, HOL,
and PL and SKL fractions F4 and F1-3 were determined by GPC (Agilent 1100, UV and RI
detectors) connected to a multiangle laser light scattering (MALLS) detector (DAWNHELEOS, Wyatt Technologies, Santa Barbara, CA, USA). To eliminate signal contributions
from lignin fluorescence, only even-numbered detectors equipped with filters were used for
molecular weight determination. For SKL, HKL, SOL, HOL, and PL, Styragel (Waters,
Milford, MA, USA) columns HR 4 and HR 2 were used, while Styragel HR 4, HR 3, and
HR 1 were used for characterization of F4 and F1-3. Experiments were run at 35 °C with
tetrahydrofuran (HPLC grade, Fisher Scientific) as the eluting solvent (0.5 mL min-1) and
the injection volume was 100 μL. The concentration of SKL, HKL, SOL, HOL, and PL were
1 mg.mL-1 (Chapter 3), while the concentration of SKL fractions F4 and F1-3 concentration
was 1, 2 or 3 mg.mL-1 as specified in Chapter 5.
2.4
Preparation of lignin-PEO solutions for electrospinning:
Electrospinning solutions were prepared as follows with some slight variations as
stated. Solutions containing only lignin and solvent without PEO were prepared simply by
adding dry lignin powder to the solvent and heating the solution in the manner outlined
below. In chapter 3, the appropriate amounts of lignin and PEO (Mw = 6 x105 g/mol) were
weighed and the dry powders were mixed using a spatula such that the weight ratio of
lignin/PEO was held constant at 99/1 or 95/5. The appropriate volume of solvent (DMF for
SKL, HKL, HOL, SOL, PL, and water for SL, LS) was then added to the mixed powders to
reach the desired total polymer concentration, which ranged from 10-50 wt%. Vials
containing solutions were then sealed tightly, vortexed for 1 minute, and heated in an oil bath
at 80oC. The solutions were again vortexed for 2 minutes after 30 and 60 minutes of heating,
and again for 3-4 minutes after 2 hours of heating and allowed cool for 15-20 minutes at
room temperature before electrospinning.
34
A slight modification of this procedure was used in preparing lignin-PEO solutions
using fractionated SKL (F4 and F1-3) in Chapters 4, 5, and 6. The powders were not mixed
with a spatula. Instead, PEO was first weighted accurately and added to the solvent, DMF.
PEO was heated in DMF at 80 °C for 10-15 minutes to dissolve it, followed by the addition
of F4 and/or F1-3 to the dilute PEO solution. The mixtures were heated at 80 °C for 2 h with
intermittent vortexing (1 – 2 min) every 30 minutes. After 2 h at 80 °C, the solutions were
allowed to incubate overnight (12-18 h) at room temperature. The solutions were then
reheated at 80 °C for 15-20 minutes, vortexed, and cooled to room temperature before
electrospinning. In Chapter 4, a range of solutions were prepared with different
concentrations of F4 and PEO as well as different molecular weights of PEO in order to study
the effect of each variable (lignin concentration, PEO concentration, and PEO molecular
weight) on the solution rheology. A list of solutions used in Chapter 4 is presented with
relevant fluid properties in Table 4.1. In Chapters 5 and 6, the PEO molecular weight and
PEO concentration were held constant at 1 x 106 g/mol and 0.2 wt%, respectively, and the
lignin concentration was varied slightly in the range 28-32 wt% in order to obtain fibres with
similar diameters but different relative amounts of the lignin fractions F4 and F1-3. The
compositions of F4/F1-3 blend solutions are presented in Table 5.1.
2.5
Electrospinning:
Electrospinning was carried out in a vertical orientation using a 1 mL syringe fitted
with a flat-tip needle as a spinneret connected to the positive terminal of a high voltage
power supply (Glassman High Voltage, Inc. High Bridge, NJ). The operating voltage was
varied from 9-15 kV in the exploratory experiments described in Chapter 3. In Chapters 4, 5,
and 6, the voltage was kept constant at 15 kV unless otherwise stated. An aluminum foil
collector was placed 20 cm below the spinneret and was connected to ground. A syringe
pump (New Era Pump Systems, Inc. Wantagh, NY) operating at a flow rate of 0.03 mL/min
supplied the polymer solution to the spinneret unless otherwise stated.
PAN nanofibres were also prepared for comparison of electrical conductivity and
Raman spectra in Chapter 6. PAN solutions were prepared by dissolving PAN powder in
35
DMF at a concentration of 10 wt% in an oil bath at 80oC for three hours with constant
stirring. Electrospinning of PAN fibres were carried out as described in the literature.298
2.6
2.6.1
Rheological characterization of electrospinning solutions:
Steady shear viscosity measurements:
In chapter 3, the shear viscosity of the lignin solutions with and without added PEO
were measured as a function of shear rate over a range of 1-2000 s-1 using an AR2000 shear
rheometer (TA Instruments, Grimsby, ON) with a cone and plate configuration (60 mm
diameter geometry with a 2o cone angle). Specific viscosity was calculated using the mean
viscosity over the shear rate range of 10-100 s-1, denoted , and the relationship sp = (–
s)/s, where s is the solvent viscosity determined by capillary viscometer at 25oC (s = 8
x10-4 Pa.s for DMF). The average shear viscosity was calculated from n=2 two measurements
at a single solution composition.
2.6.2
Small amplitude oscillatory shear rheometry:
The dynamic shear moduli of F4 and F4/PEO solutions were measured in Chapter 4
using a TA Instruments AR2000 controlled stress rheometer in stress sweep (1-1000 Pa, =
10 rad/s) and frequency sweep (oscillatory stress: 2 Pa) modes. The magnitude of the
complex viscosity |*()| = √(G’2+G”2) in the range ω = 1 – 100 rad/s was used to determine
the shear viscosity, where G’ and G” are the storage and loss moduli, respectively. In some
cases inertial effects resulted in scattered data at  approaching 100 rad/s in which case the
higher  data were not included in the calculation. Raw data from dynamic shear are shown
in the Appendix, Figure A3 and A4.
2.6.3
Capillary breakup extensional rheometry:
A capillary breakup extensional rheometer (Haake CaBER 1, Thermo Scientific,
Ottawa, ON) was used to characterize the elongational fluid properties of F4 and F4/PEO
36
solutions in Chapter 4. A vertical column of fluid was created by loading a sample between
two horizontal circular plates (diameter = 4 mm) and rapidly raising the upper plate from 2 to
11.5 mm in 50 ms with a linear stretch profile. The midpoint diameter of the fluid column
undergoing capillary thinning was measured after the upper plate came to a rest after the
initial step strain. The elongational properties were determined from the data as discussed in
the literature.299–302 Where applicable (for elastic solutions), the characteristic time scale of
tensile stress growth in uniaxial elongational flow λ (herein referred to as the relaxation time)
was obtained using the following equation to fit the portion of the thinning profile where
exponential thinning was observed:
(1)
Where Dmid(t) is the midpoint filament diameter as a function of time t, D1 is the initial
filament midpoint diameter just after cessation of the upper plate, G is the elastic modulus,
and σ is surface tension229,299–304 An example showing the calculation of l is shown in the
Appendix, Figure A5. An apparent transient elongational viscosity, ηe,app was estimated using
the equation:301
(2)
The denominator of equation 2 was an exponential equation calculated by differentiating the
exponential obtained by fitting equation 1 to the raw data from the CaBER in the region
corresponding to the elastocapillary balance, where a semi-logarithmic plot appeared to be
linear. This was typically observed in an intermediate period not including the early and late
portions of thinning as shown in the Appendix, Figure A5.
2.7
Thermostabilization of electrospun fabrics:
Oxidative thermostabilization was conducted on electrospun fabrics with dimensions
of 7.5 cm x 10 cm. The fabrics were carefully removed from the aluminum collector and
37
clamped on their edges between four L-shaped TeflonTM-wrapped glass plates. Mounting the
fabrics in this way allowed tension to be applied and maintained on the sample edges during
heating due to sample shrinkage (see Figure 5.3). The electrospun fabrics were heated in a
gas chromatography oven (Hewlett-Packard 5890 Series II) at different rates from 0.5-5
o
C/min in air to 250 oC and held isothermally for 1 hour.98 Under these heating conditions
one group of fibres (those containing only F4) remained as non-bonded fibres and will be
herein referred to as NBF, while a second group (those containing a 70/30 wt/wt mixture of
F4/F1-3) became bonded at their intersections and will be referred to as BF. Lignin fibres
thermostabilized at 250oC will be therefore referred to as BF-250 and NBF-250.
Stabilization of PAN fibres were carried out as described in the literature.298
2.8
Carbonization of electrospun fabrics:
Carbonization of the thermostabilized lignin and PAN nonwoven fabrics (Chapter 6)
was carried out in a GSL-1100X tube furnace (MTI Corp, Richmond, CA) by clamping
strips of electrospun fabric (~0.5 cm wide x 6 cm long) at each end between two stainless
steel plates and heating at 20oC/min to 250oC, followed by heating from 250oC to 600, 800,
or 1000oC at 10 oC/min for lignin and 5 oC/min for PAN298 under a N2 gas flow. The
maximum carbonization temperature (denoted Tc) was held for one hour and thereafter
allowed to cool to room temperature overnight under flowing N2 gas. The yield (relative to
the weight of stabilized fibre) of carbonization was determined immediately after removal
from the furnace by weighing the carbonized fabrics. Carbonized NBF and BF lignin-based
fibres will be referred to as NBF-600/NBF-800/NBF-1000, or BF-600/BF-800/BF-1000
corresponding to whether the fibres were bonded at their intersections (NBF for non-bonded,
BF for bonded) and the relevant Tc (600, 800, or 1000oC) used during carbonization.
38
2.9
2.9.1
Characterization of electrospun fabrics:
Optical and scanning electron microscopy of fibres obtained by electrospinning:
Fibre morphology was characterized by optical microscopy using an Olympus BX41
microscope and/or by SEM analysis (Hitachi S3000N) using gold coated samples for asspun and thermostabilized samples. No gold coating was used to image carbonized fibres.
The working distance used for SEM imaging was 10-15 cm, and the accelerating voltage
was 5 kV. Fibre and particle diameter distributions were generated from SEM images by
measuring 15-30 diameters per image using the ImageJ software package (U.S. National
Institutes of Health). Diameters are reported as the mean + standard deviation based on
measurements of 100-200 fibres.
2.9.2
Atomic force microscopy (AFM) of thermostabilized electrospun materials:
AFM images were obtained using a Veeco Instruments (Santa Barbara, CA, USA)
Mulitmode AFM using a TAP150A probe, which is mounted on a cantilever with nominal
spring constant of 5 N/m. Samples were prepared by mounting them on magnetic discs using
double-sided adhesive tape. The cantilever deflection sensitivity was calibrated on a
sapphire calibration standard and the cantilever spring constant was determined by the
thermal tune method as detailed in the instrument manual. The instrument was run in
ScanAsystTM mode with quantitative nanomechanical analysis engaged to enable
simultaneous acquisition of height vs. area and adhesion force maps over the same sample
area using a TAP150ATM probe. The applied force was fixed at 20 nN with 0.1 Hz scan rate,
30 x 30 m scan size and 512 samples/line sampling frequency. The AFM instrument
sensitivity was estimated to be on the order of ~100-200 pN under the conditions used for
imaging based on the noise of the sample traces.
39
2.9.3
Characterization of mechanical properties of electrospun fabrics:
The mechanical properties of the thermostabilized and carbonized materials were
measured as described in the literature38 using a Kawabata KES-G1 microtensile testing
system on specimens measuring 0.5 cm in width and 4 cm in length mounted on paper
sample holders. The gauge length for tensile tests was 3 cm. The samples were elongated at
an extension rate of 0.01 cm/s and the load in grams was measured as a function of time with
a load sensitivity of 200 g/V and sampling frequency of 50 Hz. Time was converted to
displacement in cm by multiplying by the extension rate. The measured load in grams was
converted to specific stress (g/tex) as calculated by the equation:38
Specific Stress (g/tex) = [Load(g)/width(mm)]/[Areal density (g/m2)]
The areal density is simply the weight of the test specimen in grams divided by the
area in m2 (length x width). The specific stress in g/tex was then converted to N/tex by
multiplying by 0.0098 and further to GPa by multiplying by the bulk density of lignin which
was assumed to be 1.35 g/cm3.305 The strain was taken as the displacement divided by the
gauge length multiplied by 100 to express it as a percentage.
2.9.4
Electrical conductivity of carbonized materials:
The DC resistance R ( of carbonized samples was measured by two-point probe
using a multimeter. Samples roughly 1.5 cm in length and 0.5 cm in width were painted at
each end with silver paint (Ted Pella, Inc. Redding, CA USA) and mounted onto clean glass
slides. Conductivity  (S/cm) was calculated based on the measured R in  and the
dimensions of the sample using the equation:
S/cm = L/(w.t.R),
Where L was taken as the distance between the two probes in cm, w was the sample
width in cm, and t was the thickness of the sample measured using a calibrated optical
microscope.
40
2.9.5
BET surface area measurements on carbonized electrospun materials:
The BET surface areas (SBET) of the carbonized materials were determined by N2
adsorption-desorption isotherms measured at 77oK using a Micromeritics ASAP 2020
analyzer. Samples were degassed at 523oK for 18 h under vacuum (500 m Hg) before being
analyzed. Five N2 uptake measurements made in the pressure range P(N2)/P(N2)0 of 0.05 0.30 were used to calculate the SBET values, where P(N2) and P(N2)0 are the equilibrium
pressure of N2 and saturation pressure of N2 at 77oK, respectively. The five ratios
P(N2)/P(N2)0 used in the analysis were 0.05, 0.11, 0.18, 0.24, and 0.30.
2.9.6
Characterization of carbonized materials by Raman spectroscopy:
Raman spectra of carbonized fabrics were recorded on a RM1000 Raman microscope
system (Renishaw, Gloucestershire U.K) equipped with a 785 nm diode laser. A total of 4
scans per sample at 1% laser power (9-10 mW) were collected using a 20X microscope
objective. The laser spot was approximated as an ellipse with major axis of 100 m and
minor axis of 7 m in order to estimate the power density on the sample as 1818 W/cm2.
Baseline correction by simple interpolation between the data points located at 900 and 1800
cm-1 was applied before curve fitting. The D-band at 1310 cm-1 was fitted with a Lorentzian
line shape and the G-band was fitted with a Breit-Wigner-Fano (BWF) line shape. ID/IG was
calculated as the ratio of the intensities (heights) of the D and G band.306,307 The G-band
position was calculated based on the BWF coupling coefficient Q and the G-band full-width
at half-maximum (FWHM) as reported in the literature.306
2.9.7
Wide angle X-ray diffraction:
Wide angle X-ray diffraction patterns of carbonized materials were obtained using a
Bruker D8 Focus diffractometer equipped with a LynxEye detector and a Cobalt source
(wavelength = 0.179 nm) over a range of 2 from 10o to 80o.
41
Chapter 3. Electrospinning of technical lignins for the production of fibrous networks
3.1
Introduction:
Lignin has relatively poor ability to form fibres by electrospinning compared to
linear, high molecular weight polymers. The addition of poly(ethylene oxide) (PEO) has been
shown to facilitate fibre formation of other biopolymers from solution by electrospinning.
These include cellulose acetate21, keratin22,23, alginate24, and chitosan25. Furthermore, it has
been demonstrated that PEO forms a miscible blend with technical lignins including kraft
and Alcell lignins26-29, and that PEO improves the ease of fibre formation by fusion
spinning7. Therefore, it was hypothesized that incorporation of PEO would facilitate
electrospinning of technical lignin solutions. In chapter 3 the effect of lignin concentration,
PEO addition and processing parameters on electrospinning of seven (7) different technical
lignins is investigated. To explore the effect of PEO addition, the effect of shear viscosity on
electrospinning behaviour was investigated for lignin solutions with or without PEO.
3.2
Electrospinning of technical lignin solutions without PEO:
Solutions of each technical lignin in DMF (SKL, HKL, HOL, SOL, PL) or water (LS
and SL) were prepared over a range of concentrations from 10-50 wt%. Unfortunately,
attempts to electrospin the various lignin solutions failed; none were capable of uniform fibre
formation, only producing electrospray. However, the SKL/DMF system did display visible
evidence of beaded fibre formation at high SKL concentration. Figure 3.1 shows SEM
images of material produced from 40 wt% (Fig 3.1a) and 50 wt% (Fig 3.1b) SKL/DMF
solutions. Increasing the concentration of SKL from 40 – 50 wt% clearly shows the transition
from electrospray (Fig 3.1a) to electrospinning with the formation of beaded fibres (Fig
3.1b). Unfortunately, further increasing the lignin concentration above 50 wt% resulted in
highly viscous solutions which produced uneven jetting and caused large droplets to be
emitted onto the collector. Nevertheless, it is apparent that the transition to fibre formation is
42
favored by increasing the polymer concentration and solution viscosity which is known to
promote fibre uniformity.40,230,231
Figure 3.1: SEM images of SKL solutions electrospun at 40 wt% (a), and 50 wt% (b). Scale bar = 100 m
It is clear that technical lignin structure and properties affect the ability to form fibres,
since one of the lignins was able to form beaded fibres and the others electrosprayed.
Interestingly, the SKL had the highest viscosity compared to the other lignins at the same
concentration dissolved in DMF. Table 3.1 shows the average steady shear viscosity for the
various lignin/DMF solutions at 30, 40 and 50 wt% concentrations.
Table 3.1: Molecular weight (n=2), polydispersity, and steady shear viscosities () of lignin solutions (n = 2) in
DMF without PEO at 30, 40, and 50 wt%. The mean + standard deviation of  at x wt% are reported as x,DMF.
Lignin
Mw
(g/mol)
Dispersity
(Mw/Mn)
30%, DMF
(mPa.s)
40%, DMF
(mPa.s)
50%, DMF
(mPa.s)
SKL
HKL
SOL
HOL
PL
3700
2500
2200
2300
2600
8.4
7.6
6.6
7.9
7.5
24 + 1.0
12 + 0.4
7 + 0.2
8 + 0.3
6 + 0.1
158 + 8.0
53 + 3.0
22 + 1.0
28 + 0.4
21 + 0.4
2341 + 58
427 + 29
94 + 1.0
178 + 4.0
127 + 2.0
43
There is a clear concentration dependence on viscosity, wherein all of the lignin
solutions increase in viscosity by an order of magnitude over the range of 30 wt% (~10 mPa
s) to 50 wt% (~100 mPa s) with the exception of SKL which increased by over 2 orders of
magnitude (~20 to ~2000 mPa s). Since the lignins are of similar molecular weights (~ 22003500 Mw), the large difference in the SKL system may be a result of the differences in the
molecular structure and intermolecular interactions as compared to the other lignins.257,294
3.3
Electrospinning of lignin with addition of PEO:
It is evident from the literature that poly(ethylene oxide) (PEO) can be used to
facilitate biopolymer electrospinning.33,35–37 In the development of lignin-based carbon fibres
it was observed that PEO levels greater than 5 wt% led to fibres fusing together during
thermal processing.98 The effect of PEO addition on lignin electrospinning was therefore
evaluated at lignin/PEO mass ratios of 99/1 and 95/5 from solutions with total concentrations
of 20-50 wt%, where concentration in wt% is expressed as % weight of (lignin + PEO) with
respect to total solution mass. Figure 3.2 shows SEM images of electrospun SKL/PEO
solutions. At a lignin/PEO ratio of 95/5, the 20 wt% solution produced beaded fibres (Figure
3.2a). Increasing the solution concentration to 25 wt% resulted in nearly uniform fibres with
a few beads (Figure 3.2b), while larger diameter uniform fibres were obtained by increasing
the concentration to 30 wt% (Figure 3.2c). Decreasing the SKL:PEO ratio to 99/1, resulted in
higher concentrations being required to obtain uniform fibres, but the concentrations were
still lower than the 50 wt% required to form beaded fibres for SKL without PEO. At the 1%
PEO content, the 30 wt% solutions formed mostly beaded fibres (Figure 3.2d), the 35 wt%
solutions formed fibres with a few beads (Figure 2e), and the 40 wt% solutions formed
uniform fibres (Figure 3.2f). Further increasing the total concentration resulted in increasing
fibre diameters, and ultimately the fibres appeared fused at their points of contact, suggesting
that solvent evaporation was incomplete at these higher concentrations.
44
Figure 3.2: SEM images of 95/5 and 99/1 SKL/PEO fibres electrospun from solutions at different
concentrations. (a) 95/5, 20 wt%, (b) 95/5, 25 wt%, (c) 95/5, 30 wt%, (d) 99/1, 30 wt%, (e) 99/1, 35 wt%, (f)
99/1, 40 wt%. Scale bar = 20 m (2000X magnification).
It is clear that the addition of PEO enables the continuous electrospinning of lignin
fibres; in the absence of PEO the 40 wt% solutions of SKL (Figure 3.1a) or any other lignin
in solution only electrosprayed. Moreover, increasing the PEO content reduced the total
polymer concentration required for fibre formation. Since lower polymer concentration is
typically correlated with reduced fibre diameter,40,230,231 we initially expected that lower
concentration might produce smaller fibres. However, 95/5 SKL/PEO fibres produced from
30 wt% solutions had similar diameters (1363 ± 234 nm) to those produced from 40 wt%
solutions of the 99/1 SKL/PEO mixture (1318 ± 251 nm). It seems that in this case increasing
the relative amount of PEO may counterbalance the tendency to reduce the fibre diameter
arising from lower total polymer concentration. This point will be addressed further in
Chapter 4.
Since a SKL/PEO mass ratio of 99/1 was sufficient for fibre formation, the effect of
PEO addition on the electrospinning of the other six technical lignins was carried out using
the 1% PEO content. For all technical lignins a clear transition from electrospray or beaded
45
fibres to uniform fibres was observed with increasing total concentration. Figure 3.3 presents
the SEM images of all of the lignin fibres electrospun from the 99/1 lignin/PEO solutions.
All of the lignin/PEO solutions were electrospinnable at the same total polymer
concentrations where the lignin solutions without PEO only electrosprayed. Interestingly, the
concentration of PEO in the 99/1 lignin/PEO electrospinning solutions was ≤0.5 wt%,
substantially lower than the minimum PEO concentration (5 wt%) required to electrospin
PEO from water or DMF. These observations suggest that interactions between PEO and
lignin influence the ability to form fibres by electrospinning.
Figure 3.3: SEM images of lignin fibres electrospun from 99/1, lignin/PEO solutions using different technical
lignins. (a) HKL 40 wt%; (b) PL 40 wt%; (c) SL 30 wt%; (d) SOL 50 wt%; (e) HOL 40%; (f) LS 30 wt%. All
scale bars = 20 m (2000X magnification).
Table 3.2 lists the fibre diameters and standard deviations of the electrospun fibres
produced from the different 99/1 lignin/PEO solutions. For almost all of the lignin/PEO
solutions, a concentration of 40 wt% was sufficient to form uniform fibres, the exception was
the SOL/PEO solution, which required a slightly higher concentration.
46
Table 3.2: Mean fibre diameters + standard deviation (n = 200) for 99/1 lignin/PEO fibres obtained from
different technical lignins
Lignin
Concentration
(wt%)
Diameter
(nm)
SKL
HKL
SOL
HOL
PL
LS
SL
40
40
50
40
40
30
30
1318 + 251
1085 + 188
1517 + 415
1135 + 171
912 + 176
1645 + 371
702 + 186
The water-based systems, LS and SL, formed uniform fibres at the lowest total
concentration of all the lignins. At 30 wt% the LS produced uniform electrospun fibres, albeit
quite large in diameter. We speculate that this is due to the higher molecular weight of
lignosulfonates as compared to other technical lignins.180,182 As expected the lower molecular
weight SL formed smaller fibres in the same solvent although with a few beads (Figure 3c).
From our experiments it was evident that while all systems formed fibres, the fibre
morphology depended on the technical lignin, as well as the operating parameters. In all of
the systems, the higher concentration solutions (DMF and H2O) required a larger collection
distance in order to allow time for fibre solidification. In some cases, average fibre diameters
differed slightly when different voltage, flow rate, and collector distance were used. The
values reported in Table 3.2, were obtained using the same operating parameters in DMF (10
kV, 0.03 mL/min, 14 cm) and water (14 kV, 0.03 mL/min, 20 cm), respectively.
It has been reported that fibre diameter scales with concentration.40,230,231 Therefore,
the relationship of fibre diameter with concentration was investigated for the various
lignin/PEO systems. Figure 3.4 shows the fibre diameter vs. concentration for the 99/1
SKL/PEO system; an essentially linear increase in fibre diameter with concentration was
observed. Fibre diameters for all the different systems were in the range of roughly 200 nm
to over 5 microns, with the largest fibres being produced from aqueous solutions at higher
concentration. Fibres with diameters around 1 micron and above were usually observed to be
relatively free of bead defects, but below 1 micron bead frequency increased with decreasing
47
diameter. Although quite large relative to other electrospun nanofibres, which can be smaller
than 100 nm in diameter,28–30 the diameters of the lignin fibres produced here by
electrospinning are roughly 25-50 times smaller than those obtained by thermal spinning.98
Moreover, they were produced from 7 different technical lignins, including one (SKL) that
was shown to have poor thermal processibility in previous work.98,294
3000
2500
Diameter (nm)
2000
1500
1000
500
0
34
36
38
40
42
44
Concentration (wt%)
Figure 3.4: Plot of fibre diameter vs. concentration for the 99/1 SKL/PEO system. Diameters are reported as
mean + standard deviation, based on 100 fibers for each solution and n=2 solutions prepared at each
concentration.
3.4
Effect of shear viscosity on fibre formation and diameter:
It is known that the morphology of electrospun fibres depends on the fluid properties
such as viscosity, relaxation time, surface tension, and conductivity.232,233 The solution
properties are functions of the polymer-solvent system, chain entanglements, and/or specific
intermolecular interactions such as hydrogen bonding240 or associative interactions.241
48
Previous investigators have used steady shear viscosity data to characterize the relationship
between polymer concentration and entanglements, and reported correlations of viscosity
with fibre diameter and electrospinning behavior.40,231,240,35,37,232,233
To investigate the scaling relationship of viscosity with concentration, a logarithmic
plot of specific viscosity (sp) vs. concentration was generated for SKL and 99/1 SKL/PEO
blends (Figure 3.5). There is a clear increase in slope at approximately 25-30 wt% for both
the SKL and SKL/PEO systems. Based on the literature, this may represent a threshold
between regimes above which interactions or entanglements between chains exert a stronger
influence on the value of sp. The intersection of two extrapolated linear regressions (R2 =
0.97-0.98) over the ranges 10-25 and 30-45% for solutions with and without PEO were
calculated to compare the relative threshold of the upturn in viscosity. The intersection for
SKL alone was 28 wt%, while the intersection for SKL:PEO was 27 wt%. The observed
difference between SKL and SKL:PEO is small or negligible, but would be consistent with a
slightly lowered entanglement threshold in the presence of PEO. Clearly the concentration
must be above this threshold to form beaded fibres for SKL with or without PEO, although a
higher concentration is required for SKL alone.
Figure 3.5: Plot of specific viscosity vs. concentration for SKL and SKL/PEO systems.
∆ = SKL:PEO, ♦ = SKL. Each point represents an average of n=2 viscosity measurements.
49
The shape and slope of the plots were very similar for SKL and SKL/PEO, although
the SKL/PEO had slightly higher values of specific viscosity. The calculated fits produced
dependencies of sp ~ c2.2 and c2.3 from 10-25 wt% and sp ~ c7.8 and c7.4 from 30-45 wt% for
SKL and SKL/PEO, respectively (see Appendix Figure A1 and A2 for fitting and
calculation). The calculated values are higher than values expected based on theory for linear
polymers in good or theta solvents in the semi-dilute range (1.3 in the unentangled and 3.74.7 in the entangled regime),308 although higher values than predicted have been
reported.37,230,240,241 Notably, such values have been observed in systems involving hydrogen
bonding240 and association complexes.37,241 Similarly, higher values can also be expected as
solvent quality decreases308,231 and also, depending on the branch chain length and
concentration, in the case of branched polymers.239 The higher observed values for the SKL
and SKL/PEO solutions are therefore consistent with the fact that technical lignins are
branched macromolecules which assume compact conformation in solution,305,309 and
participate in hydrogen bonding and associative interactions.310 Furthermore, based on the
Mark-Houwink-Sakurada parameter , which has been measured to be around 0.1 in DMF at
318 or 350 K,305 DMF is not a good or theta solvent for Kraft lignins. The addition of PEO
increases the viscosity by an average of 36% but doesn’t dramatically change the scaling
behavior of viscosity with concentration. These observations confirm that the specific
viscosity is primarily due to the lignin component, with PEO adding a relatively minor
contribution. This observation makes sense since lignin accounts for 99% of the solid content
in solution. Importantly for the present work, the plots for SKL with or without PEO are
similar while their electrospinning behaviors are different. This observation suggests shear
viscosity does not completely describe the differences with or without PEO in regards to the
ability to form fibres. For example, the difference in viscosity between the SKL solution at
50 wt% (which formed beaded fibres) and the 99/1 SKL/PEO solution at 40 wt% (which
formed uniform fibres) was roughly ~2 vs. 0.4 Pa.s. All other lignin/PEO systems displayed
similar trends, in that lower viscosity lignin/PEO solutions were capable of forming fibres
while higher viscosity lignin solutions electrosprayed.
50
To further clarify the role of shear viscosity, electrospinning was also carried out
from solutions of the same viscosity and using the same operating parameters with different
technical lignins and PEO (lignin/PEO ratio of 99/1) to investigate the effect of different
lignins on fibre diameter. At a viscosity of 8.6 x 10-2 Pa*s, PL at 43 wt% and HOL at 42 wt%
with PEO had similar diameters (1850 nm and 1700 nm, respectively). On the other hand,
there were clear differences when comparing other lignins at the same viscosity. For
example, SKL solutions at 35 wt% with PEO at the same viscosity formed fibres with
average diameters of 750 nm, while the average diameters for the HKL solutions at 39 wt%
was 1250 nm. Another difference was that the SKL fibres had a few beads while the HKL
fibres were more uniform. In addition, LS and SL solutions with PEO at the same viscosity
(30 and 40 wt%, respectively,  = 9.0 x 10-2 Pa*s) formed fibres with very different
diameters. The LS fibres were roughly 1600 nm, while the SL fibres were much larger in
diameter, and over a wide range of 5-10 microns. As mentioned previously, LS is expected to
have higher molecular weight compared to SL, and this is confirmed by the higher
concentration of the SL/PEO solution (40 wt%) required to match the viscosity of LS/PEO at
30 wt%. These observations indicate that the fibre diameter depends on the fluid properties
at a particular concentration, which depend on polymer structure and solvent. Different
technical lignins are therefore expected to have a slightly different relationship between their
fluid properties with concentration, corresponding to slightly different diameters at a given
concentration depending on the structure of the technical lignin and solvent.
These observations are consistent with the work of Yu and coworkers, who
demonstrated no correlation of the shear viscosity with the ability to form fibres from elastic
PEG-PEO aqueous solutions.229 The authors showed that elasticity as characterized by fluid
relaxation time and steady elongational viscosity was critical in the prevention of jet breakup
during electrospinning of their system. Our lignin/PEO solutions bear a resemblance to the
aqueous Boger fluids prepared by Yu and coworkers in that both contain a relatively small
concentration of higher molecular weight PEO with another oligomeric species (0.1-0.2 vs.
0.5 wt% PEO or less in our study). The oligomeric species in the work of Yu et al was PEG,
and in the present work was replaced by technical lignins, which are generally complex,
polydisperse mixtures of branched polyaromatic oligomeric and polymeric species. It is
logical based on the report of Yu et al to hypothesize that the addition of relatively high
51
molecular weight PEO to lignin in solution increases the fluid elasticity, since it was
demonstrated that small amounts of high molecular weight PEO can dramatically influence
fluid elasticity and electrospinnability without significantly altering the viscosity.229 Dilute
solutions of high molecular weight polymers in viscous solvents or in the presence of
oligomeric species are often highly elastic in elongational flow.299,300,311 Furthermore, it is
known that fluid properties in shear might not be a direct indicator of elongational fluid
properties.239 Since electrospinning is a case of strong elongational flow with high strain
rate,226 nonlinear rheological behavior (extensional thickening) would be expected to occur
with elastic fluids during electrospinning.219,222 If the addition of PEO to lignin solutions
significantly increased the fluid elasticity while only slightly affecting the viscosity, shear
viscosity measurements might not completely describe the electrospinning behavior of
lignin/PEO solutions, since these measurements provide no information on extensional
thickening. Therefore, the elongational rheology of lignin/PEO solutions will be investigated
further in Chapter 4.
3.5
Conclusion:
Seven different technical lignins were readily electrospun into fibres through the
addition of PEO (1-5 wt%). To the best of our knowledge this is the first reported system
which allows the formation of electrospun fibres from a variety of technical lignins using a
single spinneret. In the absence of PEO, none of the lignins could be processed into uniform
fibres, although beaded fibre formation was observed for the SKL system at high
concentration (>50 wt%). As with other polymer systems, a linear increase in fibre diameter
with increasing lignin concentration was observed. However, at the same concentration, the
various lignin solutions had varying viscosities and different electrospinning behavior, i.e.
fibre diameter and ability to form uniform fibres. Similar results were found using
lignin/PEO solutions with the same viscosity. Together these results suggest lignin specific
structures and intermolecular interactions are influencing solution properties and
electrospinning behavior. Further support of this was observed from the scaling exponents
calculated based on specific viscosity vs. concentration plots, which at beyond 27-28 wt%
concentration  ~ c7.4-7.8, consistent with a branched polymer participating in intermolecular
52
interactions such as hydrogen bonding or association complexes. On the other hand,
lignin/PEO solutions produced uniform fibres at viscosities much lower than those of lignin
alone that only electropsprayed, suggesting shear viscosity plays one, but not the only key
role in determining electrospinnability.
53
Chapter 4. Effect of elongational rheology on electrospinning of softwood Kraft
lignin
4.1
Introduction:
This study seeks to elucidate the rheological phenomena underlying the improved
stability of lignin/PEO solutions during electrospinning observed in previous studies, by
considering the elongational rheology of the spinning solution. It is known that polymeric
fluids undergoing large, rapid deformation typically exhibit nonlinear rheological behavior
(e.g. shear thinning, extensional thickening), and electrospinning is an example of a large
rapid deformation.218,219 While it is generally known that the rheological properties of
polymer solutions strongly affect the ability to form fibres and its resulting diameter in
electrospinning,218,219,222,225,229 the relationship between elongational rheology and the
electrospinning behavior of lignin solutions has not been reported in the literature. A deeper
understanding of the viscoelasticity of lignin and lignin-polymer blend solutions is of vital
importance to the optimization of lignin electrospinning, and should provide valuable insight
into the behavior of lignin with respect to other fibre spinning processes. It is hypothesized
that strain hardening may be a key factor governing the electrospinning of lignin/PEO
solutions. Here we report on the effect of lignin concentration, PEO concentration, and PEO
molecular weight on the rheological properties of electrospinning solutions using dynamic
shear and capillary breakup extensional rheometry (CaBER) and investigate the correlation
between rheology and fibre diameter. Softwood Kraft lignin (SKL) was selected for
rheological investigation from the previously investigated lignins due to its interesting ability
to form beaded fibres without PEO. Also, it was observed early on that CaBER was very
sensitive to inhomogeneities and/or incomplete lignin solubility in the solvent, DMF. In order
to improve the homogeneity of the electrospinning solutions for rheological characterization,
sequential extraction with organic solvents was performed on SKL as described in Section
2.2. The rheological and electrospinning experiments described in Chapter 4 were performed
using the F4 fraction.
54
Table 4.1: Fluid compositions, viscosity , relaxation time , surface tension , and corresponding fibre
diameters. - = too small to measure, *= incomplete fibre solidification during spinning, x = no fibres formed,
n/a = not applicable. n represents the number of samples used to obtain solution properties.
F4
conc.
(%)
PEO
mol.
wt.
(g/mol)
PEO
conc.
(%)
|
(mPa*s)
n=2

(ms)
n=2
25
-
0
19
-
6
0.1
24
-
32
-
6
1x10
0.2
28
-
32
-
5x106
0.1
27
~4
32
443
65
6
5x10
-
0.2
0
32
45
11
-
32
32
641
-
79
-
1x106
0.1
61
8
33
582
84
1x106
0.2
70
12
32
702
83
6
0.1
63
24
33
821
99
6
5x10
-
0.2
0
72
116
31
-
32
34
1010
-
149
-
1x106
0.1
177
20
33
895
98
6
0.2
200
24
33
1100
138
6
5x10
0.1
168
53
33
1401
231
5x106
-
0.2
0
194
448
55
-
33
33
1551
-
338
-
1x106
0.1
546
49
34
1482
229
6
1x10
0.2
523
65
34
1890
410
5x106
0.1
539
107
34
2176
494
6
5x10
-
0.2
0
586
1598
147
-
34
35
2501
420
1x106
0.1
2023
160
35
2658
440
1x106
0.2
2082
182
35
2878
502
6
0.1
2007
326
35
3261
350
6
0.2
2023
510
35
*
*
1x10
30
5x10
35
1x10
40
45
Mean
Standard
Fibre

deviation
(mN/m) Diameter
(nm)
(nm)
n=2
n=200
n=200
32
-
5x10
5x10
55
4.2
Solutions:
The solutions used for rheology and electrospinning studies are summarized in Table
4.1. F4/PEO solutions were prepared such that the range of concentrations was 25-50 wt% F4
relative to the total mass of the solution prepared in 5 wt% increments (25, 30, 35, 40, 45,
50). Four F4/PEO solutions were prepared at each F4 concentration from 25-45 wt%, where
the PEO concentration was 0.1 or 0.2 wt% and the PEO molecular weight was 1x106 or
5x106 g/mol.
4.3
Dynamic shear rheometry:
The dynamic storage (G') and loss (G") moduli were measured as a function of
oscillatory stress amplitude o ( = osin(t)) and frequency  to investigate the linear
viscoelastic regime of F4 and F4/PEO solutions. In the limit of linear viscoelasticity, all of the
solutions can be considered to be weakly elastic under shear deformation, since G" >> G' by
roughly an order of magnitude. Figure 4.1a shows representative stress sweep data for 40
wt% F4/DMF solutions with different combinations of PEO concentration and molecular
weight.
Figure 4.1: (a) Stress sweep and (b) frequency sweep data for F4 and F4/PEO solutions with F4 concentration =
40 wt%. Filled symbols are G’ (Pa), unfilled symbols are G”, and unfilled symbols connected by solid lines
represent |*()| (Pa*s). ● = F4, ♦ = PEO 0.1 wt%, 106 g/mol, ■ = PEO 0.2 wt%, 106 g/mol, ▲ = PEO 0.1 wt%,
5x106 g/mol, ▼ = PEO 0.2 wt%, 5x106 g/mol
56
In the stress sweeps shown in Figure 4.1a, G” was constant as a function of stress for
F4 without PEO, but displayed a decrease with increasing stress in F4/PEO solutions. G’
increased in F4/PEO solutions compared to F4 solutions, but G’ was always less than G”. The
stress sweeps indicate that there is a weak elastic network in the F4 solutions. This elastic
network becomes noticeably stronger with the addition of PEO. A significant increase in G’
was observed despite the small amount of PEO relative to F4: less than 0.6 - 0.8% by mass.
The increase in G’ may be due to interactions between PEO and F4, such as hydrogen
bonding192 which mutually influences molecular mobility. Regardless, it appears that the
elastic network is disrupted during shear deformation, as the value of G’ drops off above 10
Pa. From the stress sweeps, a stress of 2 Pa was considered to be within the linear
viscoelastic regime and used in the subsequent frequency sweeps.
Frequency sweep data was used to calculate the magnitude of the complex viscosity
|η*()|, shown in Table 4.1 for all of the F4 and F4/PEO solutions. The values of |η*()|
varied over nearly 3 orders of magnitude from 20 mPa*s at 25% F4 to 1600 mPa*s for 45%
F4 concentration. Figure 4.1b shows G’(), G”() and |η*()| for 40 wt% F4 solutions with
different PEO concentration and molecular weight. Interestingly, a clear dependence of G’
was observed as a function of PEO molecular weight and concentration, indicating that the
elasticity of the solutions is increased by PEO addition and depends on PEO concentration
and molecular weight. This observation was consistent with the results of stress sweeps. Here
it should be emphasized that the addition of PEO led to an increase in |η*()| of roughly 3570% when comparing F4 vs. F4/PEO solutions with the same F4 concentration. On the other
hand, solutions differing in F4 concentration by 5 wt% displayed larger differences in
|η*()|where |η*()| increased by a factor of roughly 3-4 when increasing F4 concentration
by 5 wt%. Thus, the increase in |η*()| due to increasing F4 concentration was consistently
larger than that due to PEO addition. For example, all F4/PEO solutions with F4 concentration
of 35% exhibited |η*()| of 100 – 200 mPa*s, while F4 solutions at 40% without PEO
exhibited |η*()| of 400 – 600 mPa*s. Each set of 5 solutions with a constant F4
concentration therefore represented a set within a specific range of |η*()| as shown in Table
4.1.
57
Figure 4.2: Representative thinning profiles of F4 and F4/PEO solutions. a) F4 solutions without PEO and
varying F4 conc.: = F4 40 wt%, = F4 45 wt%, = F4 50 wt% , b) F4/PEO solutions with PEO conc. = 0.2
wt%, PEO Mv = 1x106 g/mol and varying F4 conc.: = F4 30 wt%, = F4 35 wt% = F4 40 wt%, =F4 45
wt%. c) F4/PEO solutions with F4 conc. = 40 wt%, PEO Mv = 1x106 g/mol and varying PEO conc.: = PEO 0.1
wt%, = PEO 0.2 wt%. d) F4/PEO solutions with F4 conc. = 40 wt%, PEO conc. = 0.1 wt%, and varying PEO
Mv; = 1x106 g/mol, = 5x106 g/mol
4.4
Elongational rheometry by CaBER:
The viscoelastocapillary thinning process was investigated with CaBER to explore
the elongational behavior of F4 and F4/PEO solutions. Figure 4.2 displays the thinning profile
58
of different F4 and F4/PEO solutions. Time is on the abscissa with t = 0 corresponding to the
time at which the upper plate stopped moving, and Dmid(t) was then recorded by the laser
micrometer. The thinning profiles are plotted as the measured filament diameter Dmid(t)
normalized by the initial filament diameter D1 at t=0 just after the imposed step strain.
In general, F4 solutions at or below 35% concentration could not be measured using
the CaBER due to breakup occurring before the end of the initial step. It has been reported
that there is a limiting viscosity below which the breakup process cannot be accurately
measured with CaBER.302 Regardless of this limitation, a good qualitative picture of the
breakup of F4 solutions was obtained by measuring at higher F4 concentrations of 40, 45, and
50 wt%. It was observed that the filament thinning profiles were linear in time, consistent
with the behavior of a Newtonian fluid as shown in Figure 4.2a.304 The elongational viscosity
of these fluids is constant in time with a value that increases with increasing F4 concentration.
In theory, equation (2) (Section 2.3.6) could be used to obtain the apparent elongational
viscosity ηe,app, and the Trouton ratio, (ηe,app/ η0) where η0 denotes zero-shear viscosity. For
purely Newtonian fluids, ηe,app = 3η0 (Trouton ratio = 3). However, it proved rather difficult
to obtain reproducible CaBER data of F4 solutions compared to F4/PEO, which made it
difficult to obtain a consistent slope to obtain an apparent elongational viscosity. It was
nevertheless confirmed that the only reproducible profiles decreased linearly in time at 40,
45, and 50 wt%. Some artefacts were occasionally observed in the data at 45 and 50 wt%.
Anomalous artefacts can be interpreted by remembering that CaBER only measures the
filament diameter at a given plane which is presumed to capture the midpoint diameter of the
thinning fluid column. However, bulges, gravitational sagging, or undissolved particles
passing through the measuring plane appear in the data as increases in the value of Dmid(t).301
Sagging can be prevented by careful selection of experimental conditions, and observed by
photographing the entire fluid column as discussed elsewhere.302 We can not rule out the
possibility that some aggregates or inhomogeneity could exist in concentrated F4 solutions,
since Kraft lignin is known to display colloidal or associative behavior.260,312,313 These results
emphasize the importance of conducting parallel experiments in both shear and extension to
better understand the rheology of solutions used in electrospinning, as viscosities of F4
solutions are difficult to measure in extension but can readily be measured in shear.
59
The addition of PEO generally resulted in a deviation of the thinning behavior from
linear to exponential in time. Figure 4.2b, 4.2c and 4.2d show representative thinning profiles
of F4/PEO solutions when either the F4 concentration, PEO concentration, or PEO molecular
weight, respectively, were varied independently. In contrast to the F4 solutions, it was
relatively easy to obtain CaBER measurements for the F4/PEO solutions, as their elasticity
allowed them to form stable fluid columns. F4 concentrations as low as 30 wt% with PEO
produced consistent thinning behavior. Solutions with higher F4 concentration, higher PEO
concentration, and higher PEO molecular weight produced pronounced exponential thinning
behavior, longer filament lifetimes, and higher values of λ, as reported in Table 4.1.
Exponential thinning behaviour is a characteristic of elastic fluids.301,302,304 The exponential
character of the thinning process in elastic fluids is believed to originate from elastic stress
generated by uncoiling and alignment of long, linear polymer chains into an extended
conformation due to the strong character of the elongational flow.299–304,314,315 The elastic
stress in the fluid column grows to balance the increasing capillary pressure, which increases
as the fluid filament decreases in diameter. Elasticity can be recognized macroscopically in
the laboratory and is often referred to as “spinnability.” This characteristic has been observed
as improvements in the ease of fibre formation during thermal extrusion for blends of Kraft
and Alcell lignins with PEO.98 This important quality which is clearly related to the ease with
which a fibre can be drawn has not been described in terms of measurable rheological
parameters for lignin-based systems until now.
Analyzing the CaBER data in terms of apparent elongational viscosity using equation
2, it can be seen that ηe,app increases exponentially with time during exponential thinning.
Figure 4.3a shows a comparison of e,app between four different F4/PEO solutions with the
same F4 concentration (40 wt%) and different combinations of PEO concentration and
molecular weight.
60
Figure 4.3: (a) Transient elongational viscosity of F4/PEO solutions with F4 conc. = 40 wt%. (b) Semi-log plot
of thinning profiles of F4/PEO solutions with F4 conc. = 40 wt%. (c) Region of CaBER data close to filament
breakup showing data scatter at small filament diameters. = PEO 0.1 wt%, 106 g/mol, = PEO 0.2 wt%, 106
g/mol,
= PEO 0.1 wt%, 5x106 g/mol,
= PEO 0.2 wt%, 5x106 g/mol.
In order to calculate e,app equations 1 and 2 were applied to the intermediate region
of the data corresponding to elastocapillary balance.302 Figure 3a shows that the
concentration and molecular weight of PEO in solution have a strong effect on the values of
e,app even though the values of |*()| are very similar, in the range of 500-600 mPa*s
(Table 4.1). Figure 4.3a shows that the F4/PEO solution with 0.2 wt% PEO, Mv=5x106 g/mol
reaches a e,app value over 100 Pa*s during thinning, while the other solutions at F4
61
concentration of 40 wt% deviate from exponential thinning at lower e,app values. However, it
is difficult to interpret the data in terms of elongational viscosities because the steady
elongational viscosity is reached at very small filament diameters where measurement by the
CaBER micrometer is less accurate. The thinning behavior does appear to deviate from
exponential thinning near breakup (Fig. 4.3b, 4.3c) as reported elsewhere,299,300 but when
focused on the region of the data near breakup a stepwise decrease in Dmid(t) was observed
(Fig. 4.3c). This data scatter could be due to instability in the fluid column and/or resolution
limitations of the detector so it is not clear if this region can be considered representative of
the actual fluid behavior. If the expected linear decrease in diameter for the steady state is
assumed, then linear regression can be employed to extract an estimate of the steady
elongational viscosity. However, the somewhat low R2 values indicate the measurement is
not quantitative. Nevertheless, a rough estimate of the steady elongational viscosity based on
linear regression demonstrated that the Trouton ratio of F4/PEO solutions exceeds 100 in
some cases, well above the Newtonian value of 3, consistent with a significant strainhardening. We suggest based on these observations that strain hardening is key in
determining the morphology of electrospun F4/PEO fibres, as shown elsewhere in
electrospinning of other systems.218,219,222,225,229
During exponential thinning we can estimate the transient value of ηe,app using
equation 2. However, Stelter et al. pointed out that the transient elongational viscosity is not a
fluid property, and is not suitable for describing elongational flow behavior.300 On the other
hand comparing the time scales of viscoelastic stress growth, λ, provides an alternative
means to compare the elastic properties of the spinning solutions.229,299,300 The values of λ
obtained from CaBER are tabulated in Table 4.1.
The values of were dependent on F4 concentration, PEO concentration, and PEO
molecular weight. This observation is consistent with the literature reports on model systems
where increasing the solvent viscosity, polymer concentration, or polymer molecular weight
increased the value of λ.229,299–304,314,315 The observation that higher molecular weight
increased  can be explained in terms of the relative contributions of different relaxation
modes to the tensile stress in the fluid. While it is known that real fluids exhibit a spectrum of
relaxation times due to polydispersity and other molecular features, it has been shown that
capillary thinning is governed by the longest relaxation time, since the tensile stress
62
contributions due to all other modes relax away at earlier times during thinning.303 The
longest relaxation time corresponds to the unraveling of the longest chains in solution. The
fact that the values of λ increase as molecular weight increases is indicative of the slower
process of unraveling longer chains. A similar analogy applies to measurement of the longest
Zimm or Rouse relaxation time in shear, which can be related to the values measured in
elongational flow.301,302,314,315 However, it should be noted here that λ measured in this work
is not equivalent to the longest relaxation time as measured with small amplitude oscillatory
shear rheometry.
Increasing the concentrations of F4 or PEO also increased λ. The dependence of λ on
the concentration of the high molecular weight component in elastic liquids is presumably a
result from interaction of extended chains during elongation.299,300,302,314,315 Investigation of
dilute, elastic solutions by CaBER314 and imaging of droplet formation315 showed that λ is
dependent on polymer concentration well below the critical overlap concentration c* (i.e. in
the absence of equilibrium entanglements). This point may explain in part why
electrospinning could be carried out from un-entangled polymer solutions (as measured in
steady shear) in the work of Yu et al.229 The observed result from CaBER here suggests that
overlap of the relatively long, linear PEO chains occurs during elongational flow due to
unraveling and interaction between extended chains.
The clear dependence of λ on the F4 concentration in F4/PEO solutions is also
intriguing. It has been observed in published studies that λ increases with increasing solvent
viscosity.302 It was observed that these F4 solutions thin with Newtonian-like behaviour. In
this case increasing the F4 concentration may be analogous to increasing the solvent
viscosity, where PEO is the polymer, and the F4 and DMF together act as solvent for the
purpose of this discussion. However, most CaBER studies published in the literature use
ideal elastic “Boger” fluids. Typically consisting of a low concentration, high molecular
weight polymer dissolved in an oligomer of the same chemical constitution alone, or in the
presence of another co-solvent. On the other hand, it is also known that PEO forms a
miscible blend with Kraft lignins.192 In addition, it has been observed that miscibility and
compatibility affect the tendency of polymer blends containing high molecular weight
polymers to exhibit strain hardening.316,317 It is not currently known to what extent the
dependence of λ on the F4 concentration is due to specific intermolecular interactions such as
63
hydrogen bonding or to nonspecific confinement effects or hydrodynamic interactions.
However, it is reasonable to expect that specific intermolecular interactions would slow
down the relaxation of polymer chains due to overall increased friction with surrounding
chains. The compatibility of F4 and PEO may explain in part the large change in elongational
rheology in spite of the relatively low amount of PEO added. The interesting result to
emphasize from rheological experiments is that the rheology of the blend system is tunable in
that the elongational properties can be altered by changing the PEO concentration and
molecular weight while maintaining a low PEO content relative to F4. This is an important
finding for fibre spinning of lignin-based systems, since inherent variability in technical
lignins leads to considerable variation in the shear viscosity vs. concentration relationship as
discussed in Chapter 3. In order to produce consistent spinning behavior with different
technical lignins, PEO addition at different ratios of concentration and molecular weight
could presumably be used to prepare spinning solutions with consistent elongational
rheology.
4.5
Ability of F4 and F4/PEO solutions to form fibres during electrospinning:
Electrospinning experiments were carried out using the various F4 and F4/PEO
solutions to investigate the correlation between the ability to form fibres and the rheological
parameters |η*()| and λ. Representative images of the resulting fibres are shown in Figure
4.4. F4/PEO solutions were capable of forming beaded fibres at relatively low F4
concentration, as low as 25 wt% F4, while all F4 solutions below 50 wt% only electrosprayed.
Most of the F4/PEO solutions at F4 concentrations of 30 wt% and higher formed bead-free
fibres, with the exception of the 30 wt% F4 solution containing 0.1 wt% PEO (Mv = 106
g/mol), which formed beaded fibres. Interestingly, the 25 wt% F4 solutions containing 5x106
g/mol PEO at 0.2 wt% formed nearly bead-less fibre, demonstrating that higher concentration
and molecular weight of PEO could compensate the destabilizing effect of the lower F4
concentration. The dependency of spinnability and fibre morphology on PEO molecular
weight for the F4/PEO system has not been reported in the literature, although the effect of
molecular weight on both the viscoelasticity of polymer solutions and electrospinning is
well-known. These results demonstrate that at a given F4 concentration, increasing the PEO
64
molecular weight or PEO concentration above a certain threshold promotes the transition
from beaded fibre formation to uniform fibres. To connect spinnability with the rheological
characterization, we can also calculate a Deborah number, De, by dividing the measured 
values with the Rayleigh breakup time, tR.302 If we take the characteristic length scale r0 = 0.8
mm to approximate the radius of the electrified jet as reported by Yu et al.229 we see that the
transition from electrospray to beaded fibre formation corresponds to a De > 1. This result
indicates that when  exceeds the tR, breakup into droplets is suppressed by elastic stress on
the jet. In the absence of PEO, the solutions are Newtonian and the viscous stresses are
insufficient to stabilize the jet over an F4 concentration range of 25-45 wt% .
Figure 4.4: SEM images of fibres electrospun from solutions with different compositions
(F4 wt%/PEO wt%/PEO Mv g/mol) (scale bar = 10 m)
Interestingly, we were also able to obtain fibres from solutions without PEO at 50
wt% F4 using a slightly lower flow rate (0.01-0.02 mL/min) and higher applied potential (20
kV) compared to F4/PEO solutions. SEM images of the ~1200 nm diameter fibres produced
from a 50 wt% F4 solution are shown in Figure 4.5. Although we previously observed that
65
fibre formation began to occur around 50 wt% for unfractionated SKL, those fibres were not
uniform, ranging in fibre diameter from less than 100 to over 1000 nm with numerous
droplets and beads. The increased uniformity of the fibres obtained in the present work is
likely a result of the fact that the Kraft lignin sample was purified by extraction with organic
solvents, whereas the SKL in previous work was unfractionated. We suggest that the
extraction improved the electrospinnability by eliminating lower molecular weight
compounds and insoluble high molecular weight fragments,197 which could destabilize the
electrospinning jet.
Figure 4.5: SEM image of fibres electrospun from 50 wt% SKL solution without PEO. a) Purified SKL fraction
F4. b) SKL without purification. (Scale bar = 20 m)
To the best of our knowledge this is the first report of stable electrospinning of Kraft
lignin without the aid of polymer blending, or a coaxial spinneret. This result is also
interesting because, as shown in Figure 4.2a, the F4 solution at 50 wt% concentration
behaved as a Newtonian fluid in capillary thinning studies. However, elasticity is often
considered to be an essential factor underlying fibre formation.225,229,318 One explanation
could simply be that even though the fluid is Newtonian in elongational flow, the viscous
stress is sufficient to suppress instabilities leading to bead formation or breakup into droplets.
Another possibility is that SKL solutions could become elastic under the high strain rates
characteristic of electrospinning. This might be investigated by using the stable jet region of
66
electrospinning to measure elongational behavior at high strain rate.225,226 Another possible
explanation for the improved electrospinning performance is interfacial viscoelasticity, i.e.
the formation of an elastic skin on the jet as a result of solvent evaporation. Regev et al.
reported that solutions of bovine serum albumin which displayed Newtonian behavior in
CaBER experiments could form fibres depending on the conformation of BSA.318 The same
authors presented evidence to show that interfacial viscoelasticity at the jet-air interface
played an important role in stabilizing the jet during electrospinning. An important point to
emphasize is that multiple mechanisms could succeed in stabilizing the electrospinning jet. In
SKL/PEO solutions, elastic stresses provide additional stability while in the absence of PEO,
a higher SKL concentration is required to electrospin because stretched SKL solutions do not
appear to strain harden.
4.6
Correlation of relaxation time with fibre diameter:
Figure 4.6: Mean fibre diameter vs. relaxation time () for F4/PEO solutions
There was also a clear correlation between the fibre diameter and the measured values
of , plotted in Figure 4.6. The correlation between diameter and λ shows that increased
elasticity results in increased fibre diameter in electrospinning of F4/PEO solutions. The
diameter (d) vs  data could be fitted using an exponential of the form:
67
d = A+B(1-e-kwherek=0.0091051
While the physical significance of this relationship is not yet clear since it is merely
empirical, it provides a basis for further investigation of the relationship between fibre
diameter and elongational rheology. The fitting parameters A, B, and k were obtained by
estimating A and B as the approximate diameter as   0 (~350 nm) and the asymptotic
value of d (~3200 nm) obtained at high , respectively. The parameter k was guessed and a
solver program was used to obtain the best combination of the three parameters by
minimizing the sum of the squared difference between the measured and calculated values. It
should be noted that in practice, the solutions with  0 broke into droplets instead of
forming fibres, since De < . In addition, F4SKL solutions at 50 wt% formed fibres with
diameters greater than 1 micron while their elongational rheology suggested a Newtonian
behaviour ( = 0) so the extrapolated value of the fibre diameter at  = 0 does not seem to
have a clear physical meaning in this case.
It can also be seen that the standard deviation of the measured diameters (error bars in
Figure 4.6) increases with increasing λ. Part of the explanation for this can be found in the
SEM image in Figure 4.4f. As the  of the spinning solution increases, a deviation from
fibres with cylindrical cross sections was observed. Figure 4.4f shows what appears to be a
flattened and twisted fibre instead of a round cylindrical fibre observed at lower . Flattened
morphology led to a broader diameter distribution because in analyzing the SEM images flat
fibres essentially had two dimensions corresponding to a shorter and longer radial dimension.
Flattened morphology may be due to incomplete solvent evaporation from the core of the
fibre coupled with skin formation at the jet surface and collapse of the walls of the solid
sheath around the liquid core. It should be noted that even considering only fibres with
cylindrical cross sections, the width of the diameter distribution was still larger for solutions
with higher . Although all the factors influencing the width of the diameter distribution are
not clear, the elongational rheology of the spinning solution is clearly related to the mean
fibre diameter and width of the diameter distribution.
Our results emphasize the importance of striking a balance between shear viscosity
and elasticity to generate smaller diameter fibres while preventing bead formation. The
approach presented here should provide a good basis for future studies aimed at reducing the
fibre diameter and controlling the diameter distribution as well as measuring the effect of
68
fibre diameter on material properties, where precise control over the fibre diameter is needed.
Finally, it should be noted that measurements of λ using CaBER represents a rare case where
a single, rapidly measurable parameter can clearly be correlated with the processing behavior
for a lignin-based fluid undergoing a complex electro-hydrodynamical deformation.
Technical lignins have highly complex molecular structures. Structural complexity and
heterogeneity implies a general lack of predictable, reproducible processing behavior. This is
a limiting factor for the processing of lignin as a carbon fibre or other renewable material
precursor. We have shown here an example where blending relatively small amounts of PEO
can be used to overcome a lack of lignin processability, and a single rheological parameter,
, can be correlated with both the ability to form a fibre and the fibre diameter.
4.7
Conclusion:
The results obtained using CaBER illustrate clearly that F4 solutions exhibit linear
Newtonian-like behavior in capillary thinning. The addition of 0.4-0.8% PEO (relative to F4)
changed the solutions to non-Newtonian strain-hardening fluids, as indicated by exponential
thinning. λ was observed to depend on the concentrations of F4, and PEO, as well as the PEO
molecular weight in F4/PEO solutions. Solutions displaying λ above ~12 ms showed a
corresponding transition in the electrospinning behavior from beaded to bead-free fibres. The
fibre diameter increased with λ, indicating that the increased elasticity resisted thinning of the
jet, resulting in larger fibres. F4/PEO solutions with lower |η*()| produced smaller diameters
and more uniform diameter distributions. Interestingly, it was also observed that relatively
high concentration (50 wt%) F4 solutions with Newtonian elongational behavior were also
capable of forming fibres, suggesting either viscous or elastic stress can stabilize
electrospinning of lignin. To the best of our knowledge this is the first report on the
elongational fluid properties of lignin or lignin/PEO blend solutions for use in
electrospinning. Furthermore, the results of this study suggest that shear and elongational
rheometry measurements provide a good basis for further studies on controlling the
electrospinning behavior of other technical lignins.
69
Chapter 5. Preparation of moisture-responsive lignin-based materials
5.1
Introduction:
In previous chapters it was shown that electrospinning of lignin can be implemented
using PEO to control the elongational rheology of the spinning solution. Having established
an understanding of the effect of lignin concentration, PEO concentration, and PEO
molecular weight on the formation of electrospun fibres, the focus of chapters 5 and 6 will be
to study the effect of lignin structure and properties on the properties of electrospun materials
during thermal processing typical for the production of lignin-based carbon fibre (CF), as
discussed in section 1.5. Previous chapters provide a basis to prepare sub-micron diameter
fibres with consistent diameters while maintaining the amount of PEO in electrospun Kraft
lignin fibres below 1%, which should minimize the effect of PEO on the fibre material
properties. In chapter 5 and 6, electrospinning of blends of lignin fractions F1-3 and F4 into
electrospun fibres will be studied in order to study the influence of the thermal mobility of
lignin on the properties of electrospun materials. It will now be demonstrated that taking
advantage of differences in the thermal flow behavior of lignin fractions during controlled
thermal treatment of electrospun fibres is a way to generate lignin-based materials which
display interesting phase-separated morphologies resembling shape memory materials
(SMMs) (section 1.4) and moisture-dependent ability to change shape, similar to a moistureresponsive SMM. Characterization of the Kraft lignin fractions by dynamic rheology, NMR,
light scattering will also be presented in an attempt to draw correlations between lignin
structure and fibre material properties.
5.2
Electrospinning of lignin fractions F4 and F1-3:
Solvent fractionation is known to influence the molecular weight distribution and
thermal properties of lignin.197,198,202,203 Fractionation was therefore carried out in order to
obtain lignins with different properties.Excluding the small amount of ash and water-soluble
components removed prior to fractionation, F4 amounted to 34% of the unfractionated SKL,
70
while F1-3 was 45%. The remainder was 21% of an insoluble fraction, which was also only
partially soluble in DMF, the solvent used for electrospinning. This material was not
characterized further but is expected to consist of high molecular weight lignin as well as
carbohydrates.49,50 Excluding this insoluble high molecular weight fraction resulted in more
stable electrospinning process. The improved stability was beneficial for obtaining good
quality nonwoven fabrics (i.e. minimal spraying and droplet formation occurred during
spinning). A photograph of the as-spun fabric is shown in Figure 5.1.
Figure 5.1: Photograph of electrospun softwood Kraft lignin nonwoven fabric after electrospinning
Both the F4 and F1-3 SKL fractions readily formed nanofibres by electrospinning;
Figure 5.2 shows SEM images of F1-3 (Fig. 5.2a) and F4 (Fig. 5.2b) electrospun nanofibres
respectively.
71
Figure 5.2: SEM images of electrospun fibres obtained from solutions containing (a) 32 wt% F1-3, 0.2 wt%
PEO and (b) 32 wt% F4, 0.2 wt% PEO. Scale bar = 50 m.
In order to obtain good quality nonwoven fabrics consisting of uniform fibres free of
beads and spray defects, the concentration of PEO in solution was held constant at 0.2 wt%,
(relative to the sum of the weights of the lignin and solvent), which was selected based on
rheology/electrospinning studies from previous work. The total lignin concentration (F4+F1-3)
was varied slightly in order to obtain uniform fibres from solutions containing different
amounts of the fractions. F4 could be electrospun at slightly lower concentration compared to
F1-3 due to the higher shear viscosity of F4 solutions relative to F1-3. F4 formed uniform fibres
at 28 wt% F4, 0.2 wt% PEO while F1-3 formed fibres at a concentration of 32 wt% with 0.2
wt% PEO. The total lignin concentration in F4/F1-3 blend solutions was therefore adjusted
slightly higher than 28 wt% in order to accommodate the slight destabilizing effect of the
lower F1-3 viscosity, while maintaining the PEO content in the as-spun fibres at less than 1%
relative to lignin (0.2/32.2 – 0.2/28.2 = 0.62-0.71% PEO). Table 5.1 summarizes the
electrospinning solution compositions used.
72
Table 5.1: Spinning solution compositions, thermostabilization heating rates, and resulting morphology after
heating electrospun F4/F1-3 fibres
SKL conc.
(%, F4+F13)
PEO
conc.
(%)
F4/F13
ratio
(w/w)
Stabilization
heating rate
(oC/min)
Morphology*
28
28
30
30
32
32
32
32
32
32
32
32
32
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
100/0
100/0
70/30
60/40
50/50
50/50
50/50
50/50
50/50
40/60
30/70
30/70
0/100
0.5
5
5
5
0.5
1
2
3
5
5
0.5
5
0.5
fibre
fibre
bonded nonwoven
porous film
fibre
fibre
bonded nonwoven
porous film
smooth film
smooth film
porous film
smooth film
smooth film
* morphology obtained after stabilization
5.3
Thermostabilization of nonwoven fabrics:
As-spun fabrics containing different weight ratios of F4 and F1-3 were heated in air at
different heating rates up to 250oC and held at this temperature for one hour. After
thermostabilization the fabrics were darker in colour and exhibited some shrinking. Figure
5.3 shows the appearance of the electrospun fabric after thermostabilization. The slightly
darker areas around the edge of the fabric are the edges that were held clamped during
heating. The slight inward contraction at the fabric corners shows the shrinking that occurred
during thermostabilization.
73
Figure 5.3: Photograph of electrospun lignin fabric after thermostabilization at 250 oC in air. Unfused fibres are
shown.
When heating fibres containing different ratios of F4/F1-3 at 5 oC/min to 250 oC,
different fusion behaviors were observed depending on the weight ratios of F4/F1-3. SEM
images are shown in Figure 5.4 to illustrate the different morphologies resulting from heating
the materials at 5 oC/min.
74
Figure 5.4: SEM images of electrospun fibres containing different ratios of F4/F1-3 after thermostabilization at 5
o
C/min to 250oC, in air. (a) F4/F1-3 = 100/0, (b) 70/30, (c) 60/40, d) 50/50.
Scale bar = 10 m.
In general, four different morphologies were obtained depending on the weight ratio
of F4/F1-3 and the heating rate, which we refer to as fibre (Fig. 5.4a), bonded nonwoven (Fig.
5.4b), porous film (Fig. 5.4c), and smooth film (Fig. 5.4d). A higher relative amount of F1-3
resulted in more softening and flow of the material during heating. For example, at a constant
heating rate of 5 oC/min, an F4/F1-3 ratio of 70/30 resulted in a bonded nonwoven
morphology, while 60/40 formed a porous film and 50/50 formed a smooth film.
Furthermore, the applied heating rate also affected the degree to which the fibres fused
75
during heating. It was observed that reducing the heating rate could result in a different
morphology at a constant F4/F1-3 ratio. A 50/50 F4/F1-3 fabric, for example, fused into a
smooth film morphology (Fig. 5.4d) at 5 oC/min but could form any of the aforementioned
morphologies when heated at slower rates. The resulting morphologies for different
combinations of F4/F1-3 ratios and heating rate are also shown in Table 5.1.
Fibres consisting of only F4 maintained their fibre form during heating over the entire
heating rate range studied (Fig. 5.4a). An interesting point to note is that 5 oC/min can be
considered a relatively fast heating rate for stabilization of lignin fibres. It has been reported
that thermally extruded hardwood Kraft lignin fibres must be heated at rates below 1 oC/min
in order to prevent fibre fusion during thermostabilization.98,99 During thermostabilization
lignin undergoes two thermally induced mechanisms: (1) mobility increase due to
temperature-induced polymer relaxations and (2) mobility reduction due to high-temperature
crosslinking. A faster heating rate favors the relaxation over crosslinking; consequently
higher flow occurs forming fused fibres. Conversely, a slower heating rate favors
crosslinking before relaxation, causing progressively increasing Tg and lower polymer flow,
resulting in less fibre fusion.319 Fibre fusion has normally been considered undesirable in
previous studies and heating rates were selected slow enough for preventing fusion.
Therefore it was interesting that F4 displays no fusion even when heated at high heating rates.
This observation suggests highly restricted mobility of F4 (as will be shown later with
dynamic rheology), causing little or no flow during thermostabilization. In contrast, F1-3
fibres could not maintain fibre form even when heated at 0.5 oC/min, which was the rate
previously used for stabilization of hardwood Kraft lignin fibres.98
5.4
Moisture-responsiveness, shape change, and shape recovery:
When the content of F1-3 relative to F4 in the fibres was increased, a very interesting
behavior was observed in the thermostabilized electrospun materials: reversible, moistureresponsive shape change. To illustrate the behavior of the material, frames from a video of a
material placed on a moistened paper surface are shown in Figure 5.5.
76
Figure 5.5: Thermostabilized (5oC/min) 50/50 F4/F1-3 film placed on moist paper (a-d), then moved to dry paper
(e-h)
The underlying surface is a piece of paper with a drop of water spread onto it and a
50/50 film heated at 5 oC/min (smooth film, Fig. 5.4d) is placed on the wet spot, allowed to
change shape, and then moved to a dry spot on the paper. As seen in Figure 5.5, the material
appeared to choose a curling axis such that two edges pushed away from the moist surface.
This choice of curling axis may be related to the tension imposed on the material during
heating due to shrinking while the film edges were held clamped. The material began to
change shape immediately upon exposure to water and reached a maximum shape change
after 30-60 seconds. Most of the change occurred over the first 10 seconds, shown in the top
four images of Figure 5.5 (a-d). When the material was removed from the moist surface and
moved to a dry spot (Fig. 5.5e-h), it immediately began to return to its original shape. Shape
recovery was a slightly slower process, again occurring more quickly initially but requiring
60-120 seconds to completely regain the original shape. The material could also regain its
shape or curl the other direction by flipping it over so that the other face came into contact
with moisture. Interestingly, moisture-responsive behavior could also be activated by placing
it in the bare palm of the hand, breathing onto one face, placing the film on the surface of
liquid water, or even above but not in contact with liquid water. Weighing the material in the
uncurled vs. curled state provided a means to approximate the sensitivity of the material. It
77
was determined that ~0.5 % moisture uptake onto one face of the materials was sufficient to
initiate the curling response. The behavior is slightly different compared to other moisturesensitive shape memory polymers, many of which are thermally-responsive SMMs that are
programmed by heating and then can be immersed completely in water to activate shape
recovery, or recover their shape gradually after exposure to humid conditions for a prolonged
period.65,67,69–71,73–75 The ability to rapidly activate changes in shape upon exposure to
humidity and the relatively rapid recovery is an interesting aspect of these materials.
The ability to display moisture-responsive behavior was related to both the
morphology and the ratio of F4/F1-3. The materials displaying bonded nonwoven, porous film,
and smooth film morphology and containing F1-3 contents of 40-70% changed shape in
response to moisture and regained their original shape when the moisture stimulus was
removed. Physical bonding of the fibres caused by the flow of the F1-3 fraction upon heating
was apparently key to inducing the interesting behavior in these materials. If no F1-3 was
included in the formulation or if the blended materials were heated such that they would
maintain un-fused fibre form, the materials did not show any moisture-responsiveness or
shape change. Heating slowly allowed too much time for chemical cross-linking to occur
prior to softening in order to allow the formation of an inter-bonded physical network, and no
moisture-responsiveness was observed.
Another interesting observation was that blending of the fractions was sufficient but
not necessary for inducing moisture-responsiveness. Moisture-responsive materials could
also be obtained by stacking a fabric consisting of F1-3 fibres on top of an F4 fabric and
heating them together. As observed previously, the F4 fibres maintained fibre form during
heating and the F1-3 softened and flowed even at slow heating rate of 0.5 oC/min, creating a
smooth layer of F1-3 coating one surface of the F4 fabric. When this stacked material was
placed with the F4 surface in contact with moisture, it curled away from the surface. A
difference compared to the blended systems was that the material did not fully regain its
original shape when the stimulus was removed in this case but remained slightly curled.
Also, the material only curled in one direction. Placing the material with the F1-3 surface in
contact with moisture did not induce any change in shape. However, if an F1-3 fabric was
placed between two F4 fabrics and heated, the material became capable of curling both ways
depending on which face was placed in contact with water. These F4/F1-3/F4 sandwich-type
78
materials were also capable of recovering their original shape. These observations show that
each fraction plays a distinct role in the material behavior and that the formation of an
interfacial layer and physical network connecting the fractions was essential to induce shape
change capability. SMMs are also characterized by distinct, yet intermixed phases which take
on different roles (i.e. switching and remembering functions). It should be noted that it is not
clear whether the materials are capable of memorizing different “programmed” shapes,
therefore it is not completely accurate to describe this material as a SMM. However, the
materials are clearly stimuli-responsive, capable of changing their shape, and fully
recovering their original shape upon drying and retain their ability to do so repeatedly over
many cycles.
5.5
AFM imaging of moisture-responsive films:
In an effort to draw correlations between typical SMMs and our moisture-responsive
films, we investigated the possibility that different lignin fractions could form phaseseparated systems. Atomic force microscopy proved to be very effective in revealing the
presence of different phases in our moisture-responsive films. A 3-d height image and
corresponding 2-d adhesion force map of a 50/50 F4/F1-3 blend material heated at 5 oC/min is
shown in Figure 5.6 (produced under the same conditions as the material shown by SEM in
Figure 5.4d and Figure 5.5). The height image (Fig. 5.6a) shows that even though the
material appears smooth in SEM (Fig. 5.4d), there are still traces of the electrospun fibres
and the spaces between the fibres. The height image shows that some remnant of the network
of electrospun fibre remains after heating.
79
Figure 5.6: (a) 3-dimensional AFM height image and (b) corresponding adhesion force map of a 50/50 F4/F1-3
moisture-sensitive film heated at 5 oC/min. The size of the imaged area was 30 x 30 m
Figure 5.6b shows the same sample area as the image in Fig. 5.6a but instead of the
typical height image, the adhesion force between the sample and the AFM tip is mapped as a
function of position. Brighter areas correspond to higher adhesion forces (larger force was
required to remove the tip from the surface). Comparison of sections of adhesion force maps
of the 50/50 blend material and a surface composed of only F1-3 showed that the regions with
higher adhesion correspond to the F1-3 fraction. Examples of these adhesion sections are
shown in Figure 5.7. The top portion of Figure 5.7 shows the variation of adhesion force
along a 30 micron section of the image of a moisture responsive film with a 50/50 ratio of
F4/F1-3. The variation is roughly 8-9 nN judging from the difference between the amplitude of
the positive and negative peaks, which is nearly half the applied force of 20 nN. In contrast,
the section of the adhesion map of the F1-3 surface shows an overall variation of less than 1
nN.
80
Figure 5.7: Adhesion force sections from AFM images on 50/50 F4 blend materials (top) and a film containing
only F1-3.
One difficulty in interpreting the images was that the effect of sample roughness on
the adhesion map was not completely clear since it was difficult to control the roughness
while maintaining a constant F4/F1-3 ratio and moisture responsive behaviour. The images
appear to show that the adhesion is affected by the sample roughness/contour, but the phases
do not strictly correspond to the height contours. Figure 5.8 shows another AFM adhesion
force map superimposed on the corresponding height image.
81
Figure 5.8: Adhesion force superimposed on a height image of a moisture responsive film with F 4/F1-3 ratio of
50/50.
It is possible that some artefacts are present in the image of Figure 5.8. For example,
the adhesion force seems to reach a higher value following the tallest features on the image.
More detailed analysis on surfaces with controlled roughness might reveal additional
information on the contributions of different factors to the adhesion force measurement.
Nevertheless, it seems that AFM was able to clearly demonstrate that two co-continuous
phases were present in the moisture-responsive F4/F1-3 films, and could distinguish between
the phases in terms of the corresponding lignin fractions. AFM showed that upon heating the
blended materials, the differences in thermal mobility and presumably some degree of
incompatibility result in a phase-separated system.
The results of AFM reflect a similarity between the lignin-based films reported here
and SMMs. Polymer SMMs such as synthetic copolymers and blends of synthetic polymers
82
are characterized by the presence of two distinct phases or segments, which contribute
different aspects of stimuli-responsiveness, shape change, and shape recovery.80–82 One of the
phases is responsible for stimuli-responsiveness by acting as a switch which can be flipped
by changing its phase, conformation, or mobility. For example, a temperature-sensitive
material may recover its permanent shape upon heating through the Tg of the switching
phase. The other phase remains relatively immobile upon application of the stimulus and
allows the material to “remember” its original shape. Our experiments suggest that the
function of the F1-3 fraction is that of the moisture-sensitive “switch” and the function of the
F4 fraction is to remain immobile and allow the material to remember the original shape of
the film. To generate a better understanding of the system and the mechanism underlying
moisture-responsiveness, lignin fractions were characterized by dynamic torsional rheometry,
NMR, and GPC-MALLS.
5.6
Dynamic rheology of lignin fractions:
Figure 5.9 shows storage (G) and loss (G) moduli and tan δ for the lignin fractions
as a function of temperature.
83
Figure 5.9: Thermorheological responses of lignin F1-3 and F4 fractions. Average (n = 3) first heat storage (G’)
modulus (Top), and tan δ (Bottom) are presented.
F1-3 and F4 showed significantly different thermo-rheological responses. G drastically
decreased when F1-3 specimens were heated beyond their glass-rubber transition temperature
(Tg); the G decreased more than 4 decades across the Tg. The average Tg, as defined by the
peak tan δ, is 152±2 oC for F1-3. In contrast to F1-3, the F4 fraction showed a very small
decrease in shear moduli as a function of increasing temperature. The tan  profiles for F4
showed a maxima around 230 oC. The difference between the tan intensities and the Tg
from the two fractions is also noteworthy; significantly higher intensity for F1-3 indicates that
a higher amount of polymeric chains are involved in the segmental relaxation and
significantly lower mobility is associated with F4 relaxation. These results were consistent
with the observation that F4 fibres displayed no fusion during thermostabilization while F1-3
fibres could not maintain their fibre form due to softening and flow.
Dynamic rheology showed that fractionation by solvent extraction produced two
lignin fractions with different thermal softening behavior, in agreement with previously
84
reported results for solvent fractionation.198 It should be noted, however, that other
fractionation techniques can also produce lignin fractions with distinct thermal behaviour.
Nordstrom and coworkers have reported that ultrafiltration of Kraft lignin can be used to
obtain fractions with different thermal softening behavior. They showed that the portion of
Kraft lignin which passed through a 15 kD ceramic membrane could soften to a sufficient
extent in order to obtain fibres by melt spinning from either softwoods or hardwoods.320
These researchers showed in the same publication that permeate fractions from hardwood
could be blended with unfractionated softwood Kraft lignin in order to enable fibre formation
by melt spinning.
Dynamic rheology was further employed to study the balance between softening and
cross-linking underlying moisture-responsiveness by measuring the thermo-rheological
properties of electrospun fabrics consisting of 50/50 (F4/F1-3) blend fibres at different heating
rates. Figure 5.10 shows G and tan  for 50/50 fabrics heated at 1, 2 and 3 oC/min under
dynamic shear.
85
Figure 5.10: Dynamic rheology of electrospun fabrics of F1-3 and F4 blend (50/50). Effects of heating rates on
storage modulus (Top) and tan  (Bottom) are presented.
Lignin fabrics underwent significant softening at a temperature range of 180 to 200
C, as observed by G profiles. Beyond 200 C, the materials showed temperature-induced
crosslinking and the G increased. However, the heating rate influenced the degree of
softening; lower heating rate caused lower softening, indicating lower heating rate allowed
more crosslinking to occur by the time a given temperature was reached. Effects of heating
rate were also observed in the tan  profiles. At 3 C/min heating rate, two major peaks were
observed, one at 180 C and another at 210 C. These two peaks were attributed to lignin
softening (Tg) and high-temperature crosslinking, respectively. However, at lower heating
rates, these two mechanisms are not resolved in the tan  profile; softening and crosslinking
86
occurred relatively more concurrently, with the tan  profiles showing one broad peak.
Additionally, for the 3 C/min heating rate a weak softening was observed near 120 C,
which was less intense at 2 C/min and non-recognizable at 1 C/min. This can be attributed
to the residual solvent (c.a. 4-5 %) from the electrospinning process. With a slower heating
rate solvent has more time to gradually evaporate from the fabric, making the softening nondiscernible.
As mentioned before, 1 oC/min was slow enough to maintain a fibre form during
thermostabilization (Table 5.1) and rheology of the as-spun fabric carried out at the same
heating rate showed considerably less softening (Fig. 5.10). When heated at 3 oC/min, the
material fused into a porous film under thermostabilization conditions and when it was
heated at the same rate under dynamic shear a pronounced softening was observed. This
result suggests that slower heating allowed the material to become more chemically crosslinked by the time a given temperature was reached so that insufficient softening occurred to
allow flow and physical bonding of the fibres. At a heating rate of 1 oC/min, SEM showed
that the 50/50 fabrics do not fuse, and the materials did not display moisture-responsiveness.
On the other hand, when heated at 3 oC/min, the same 50/50 material flows, the fibres fused,
and the material became a porous film, and displayed moisture-responsiveness and shape
change capability. The results of dynamic rheology suggest that significant flow of F1-3 and
formation of an interconnected physical network is critical in inducing moisture-sensitive
shape change capability.
5.7
NMR characterization of lignin fractions:
NMR was used to characterize the chemical structure of the lignin fractions to better
understand the correlation between structure and thermal softening. 1H-NMR of acetylated
fractions was used to determine the amounts of phenolic and aliphatic hydroxyl groups.
Figure 5.11 shows the acetoxyl region of the 1H-NMR of acetylated F4 and F1-3 SKL; the
peaks at 2.5-2.2 ppm and 2.2-1.7 ppm corresponding to acetylated phenolic and aliphatic
hydroxyl groups, respectively.
87
Figure 5.11: 1H-NMR of acetylated F4 (top) and F1-3 (bottom) from 2.6-1.7 ppm, showing the peaks
corresponding to acetylated phenolic (2.5-2.2 ppm) and aliphatic (2.2-2.0 ppm) hydroxyl groups
The total number of hydroxyl (-OH) groups was similar for the two fractions (7.0 vs
6.7 mmol -OH/g lignin for F4 and F1-3, respectively), but a clear difference in the relative
amounts of aliphatic and phenolic -OH groups was observed. F4 contained higher aliphatic –
OH (4.2 vs 3.0 mmol/g), but less phenolic –OH (2.8 vs 3.7 mmol/g) than F1-3.Higher
phenolic -OH content in the F1-3 fraction is consistent with a greater extent of cleavage of O-4 ethers during Kraft pulping. The differences in the relative amounts of aliphatic vs.
phenolic –OH groups is significant because the strengths of hydrogen bonding for different
hydroxyl groups are different, which in turn affects polymer physical properties. FT-IR
studies on lignin model compounds have shown that aliphatic hydroxyl groups form stronger
hydrogen bonds and are more likely to form inter- as opposed to intra- molecular hydrogen
bonds.257 The higher content of aliphatic –OH groups in the F4 fraction therefore suggests
that there should be stronger intermolecular hydrogen bonding in this fraction, in agreement
with the low thermal mobility of F4 observed during thermostabilization and in dynamic
rheology.
88
Figure 5.12: 13C-NMR of acetylated F4 (top) and F1-3 (bottom) in the region 172-167 ppm, corresponding to
carbonyl carbons of acetyl groups
13
C-NMR was performed on acetylated samples to further investigate the differences
in types of hydroxyl groups between the fractions (Fig. 5.12).59 In the 13C spectrum of
acetylated lignin, the region from 172-169.6 ppm corresponds to acetylated primary aliphatic
-OH groups while the peak corresponding to secondary aliphatic -OH groups is located at
169.6-168.6 ppm.59 Setting the integral from 165-100 ppm due to aromatic 13C to 600, the
ratio of the peak integrations for primary/secondary aliphatic groups for F4 was 1.00, while
that for F1-3 was 0.64. These results show that F4 contains roughly equal amounts of primary
vs. secondary aliphatic –OH groups while F1-3 contains more secondary aliphatic –OH
groups compared to primary aliphatic –OH groups. These differences can also be expected to
have an effect on the physical properties of the fractions because the strength of hydrogen
bonding for primary vs. secondary alcohols is different. Uraki et al. recently published
studies on lignin -O-4 model polymers showing that the presence of -OH (primary) groups
strongly reduced the thermal mobility of the -O-4 model oligomers.321 Oligomers
containing both secondary -OH and primary -OH groups were shown to be infusible (did
not soften), while oligomers with only secondary –OH groups at the -position were fusible.
The higher relative proportion of primary aliphatic hydroxyl groups should also therefore
contribute to stronger intermolecular hydrogen bonding in F4 compared to F1-3.
The lignin fractions were also characterized in a non-acetylated state by 2-D NMR
(Heteronuclear Single Quantum Coherence – HSQC) to investigate differences in inter-unit
89
linkages. The oxygenated aliphatic region of the HSQC spectra are shown in Figure 5.13,
with 13C-1H correlations corresponding to -O-4, -’, and -5 linkages labeled.
90
F4
F1-3
Figure 5.13: Oxygenated aliphatic region (1H: 2.7-6.5 ppm, 13C: 48-95 ppm) of HSQC spectra of F4 (top) and
F1-3 (bottom). Unassigned peaks are traced in black.
91
Based on a comparison of the HSQC spectra, it was clear that both fractions showed
peaks corresponding to -O-4, -’, and -5 linkages, but since the experiments were not
quantitative, it was not possible to compare the relative amounts of linkages from HSQC.
Some differences were observed in unassigned peaks in the aliphatic and aromatic regions,
shown in the Appendix Figure A6, A7, A8, and A9.
Figure 5.14: 13C-NMR of F4 (top) and F1-3 (bottom) (a) etherified C-4 in guaiacyl units (b) -O-4,13C ’, 5, 13CcC-O-4 C’; d) C-O-4, ; (e) methoxyl.
Several areas of the quantitative 13C spectra of non-acetylated fractions (Fig. 5.14)
were integrated with respect to the aromatic region (area from 165-100 ppm = 600) to
compare the structures of the fractions.297,322,323 13C-NMR showed that F4 contains a higher
amount of the typical linkages found in lignin, -O-4, -, and -5, based on the peaks in
the region 90-58 ppm,297,322,323 and more methoxyl (OMe) groups (strong peak at 58-54 ppm)
per 100 aromatic rings compared to F1-3 (93 OMe for F4 vs. 74 OMe for F1-3 per 100 aromatic
92
rings). The integrations of different regions of the quantitative 13C-NMR spectra are given in
Table 5.2.
Table 5.2: Integration of quantitative 13C-NMR spectra of F1-3 and F4. The area of the aromatic region (162102 ppm) was set to 600. The reported values are expressed as quantities per 100 aromatic rings. Ar = aromatic,
Alk = alkyl.
Chemical
shift (ppm)
193-191
182-180
175-168
168-166
157-151
144.5-142.5
162-142
142-125
125-102
90-77
77-65
65-58
58-54
54-52
Assignment
Ar-CHO
spirodienone, quinone
aliphatic COOR
conjugated COOR
C3 in 5-5', C3 and
C5 in 4-O-5
C3 in -5, C4 in 5-5'
Ar-C-O
Ar-C-C
Ar-C-H
Alk-O-Ar, C-O Alk
C-O-Alk, secondary OH
Primary OH
OCH3
C in -5 and -'
Integration Integration
(F1-3)
(F4)
3.1
1.3
9.6
1.1
0
0.7
5.9
0.6
12.7
22.0
19.9
166.1
174.1
259.3
38.6
14.5
12.9
73.7
3.4
24.8
185.9
181.3
232.6
129.1
48.9
37.5
92.3
8.8
The 13C spectra also showed a higher proportion of etherified aromatic C-4 of
guaiacyl units (149.8 ppm)322 in F4.This indicates that F4 contains ether linkages that are
more resistant to degradation during Kraft delignification, such as those connected to
condensed 5-5’ groups. F4 contains a higher proportion of etherified 5-5’ type structures as
indicated by the higher intensity of the 157-151 ppm region, taking into account the amounts
of conjugated COOR structures.297 The carbonyl region (200-165 ppm) of the 13C-NMR
spectra of non-acetylated samples also showed differences between the fractions. The
spectrum of non-acetylated F1-3 shows the clear presence of aldehyde (193-191 ppm), and
aliphatic and conjugated COOR groups (175-166ppm), while F4 shows only a weak, broad
93
band in the region (182-170 ppm), which was also observed in F1-3. The results of 13C-NMR
show that the fractions are somewhat different in chemical structure with respect to each
other, with F4 retaining more interunit linkages typically found in lignin.
5.8
Characterization of molecular weight by GPC-MALLS:
GPC-MALLS experiments were carried out in order to compare the molecular weight
distribution (MWD) of F4 and F1-3. Figure 5.15 shows the MWD obtained by light scattering
for the acetylated lignin fractions dissolved in THF. From light scattering, the Mw of
acetylated F1-3 was 7.1 x 103 g/mol, with polydispersity index (PDI) of 1.95. Because the
light scattering signal-to-noise was very low for acetylated F1-3 at a concentration of 1 and 2
mg.mL-1, the molecular weight was calculated from the data at 3 mg.mL-1. The Mw of F4 was
determined to be 3.8 x 104 g/mol with a PDI of 1.56 measured at 1 mg.mL-1. The results of
GPC-MALLS showed that the F4 fraction has higher molecular weight and lower
polydispersity compared to F1-3. Higher molecular weight also likely contributed to the
reduced mobility of F4 compared to F1-3.
94
Figure 5.15: Light scattering data showing elution curves obtained from GPC-MALLS for acetylated fractions.
Black: F1-3, 3 mg.mL-1, red: F4, 1 mg.mL-1, green: F4, 2 mg.mL-1, blue: F4 3 mg.mL-1
In addition, concentration dependence of the apparent molecular weight revealed
further information regarding molecular interactions within F1-3 and F4. Concentration
showed significant effects on the acetylated F4 elution profiles, but not those of F1-3.
Acetylated F4 produced a peak around 20 mL elution volume and a shoulder at 22 mL at a
concentration of 1 mg.mL-1 (red curve in Fig. 5.15). At higher concentrations (2 and 3
mg.mL-1, green and blue in Fig. 5.15, respectively), the shape of the elution curve for F4
became increasingly multimodal and the apparent molecular weight shifted to higher values,
as indicated by the shift of the peak eluting at 20 mL to ~19 mL and the appearance of a new
peak eluting around 17-18 mL. Acetylated F1-3 did not show concentration-dependence (data
not shown) of the shape of the elution curve, producing only one peak at 3 mg.mL-1 (Fig.
5.15, black curve). Multimodal elution curves have been shown in numerous works to be due
to lignin association.258–260,312,313,324 Gosselink et al. also showed bimodal elution curves for
solvent fractionated Indulin-AT measured in aqueous NaOH,202 but association was not
95
discussed in their study. Morck et al. also mentioned unexplained high molecular weight
“tails” in their GPC data of acetylated lignin fractions in THF.197 The mechanism of lignin
association appears to depend on several factors and is still not completely understood.
Hydrogen bonding313 and non-bonded orbital interactions between aromatic rings258–260 have
both been suggested as driving forces for lignin association. The relative contribution of
different driving forces depends strongly on the lignin structure259 and the solvent system.
Since acetylation disrupts the tendency of lignin to form inter- and intramolecular hydrogen
bonds, its contribution to the association observed in GPC-MALLS experiments should be
diminished. Therefore, the association of F4 observed in light scattering is likely a reflection
of increased non-bonded orbital interactions between aromatic rings as suggested by
Sarkanen260 and recently studied by Deng et al.258 The results of GPC-MALLS
characterization show that both the molecular weights and the types of intermolecular
interactions expected to govern the physical properties of the fractions are different. An
increased degree of non-bonded orbital interactions between aromatic rings is another
potential explanation for the reduced thermal mobility of F4, and provides an important
perspective for understanding the mechanism of moisture-responsiveness.
The mechanism of moisture-induced shape recovery in SMMs has been attributed to
disruption of hydrogen bonds by adsorbed or “bound” water.65,66,69,70,72–75 Both lignin
fractions should display intra- and intermolecular hydrogen bonding as both contain hydroxyl
groups and ethers, and oxidative thermostabilization is expected to result in the formation of
carbonyl groups, which can also participate in hydrogen bonds.257,319 It is reasonable to
suggest that moisture can interact with these polar groups. It is also known that the thermal
mobility of lignin is strongly affected by water.22,193 Therefore the mobility of F1-3 might be
increased when hydrogen bonds are disrupted. F4 also has polar groups that can interact with
water, but since the intermolecular hydrogen bonding is stronger and the molecular weight is
higher, the network of hydrogen bonds might not be disrupted enough to increase its
mobility. A smaller fraction of disrupted hydrogen bonds could play a role in “remembering”
the original shape of the film by storing elastic energy. F4 also displayed pronounced
association under conditions where hydrogen bonding could not explain the association,
suggesting that interactions between aromatic rings also play an important role in
determining the physical properties of F4. It is reasonable to suggest that these hydrophobic
96
interactions should be much less susceptible to being disrupted by the presence of water
compared to hydrogen bonds and could therefore also contribute to the reversibility of the
shape change. F4 can act as an immobile segment due to a stronger network of intermolecular
interactions which can store elastic energy and restore the films to an uncurled state upon
drying. Since an interface between the fractions also appears to be important, the strength of
the interactions between the fractions as well as within may also be involved in moistureresponsiveness.
5.9
Conclusion:
It has been shown that solvent fractionation can be used to obtain distinct fractions
from commercially available Kraft lignin such that the different structure and properties of
the fractions can be exploited in the preparation of novel electrospun, moisture-responsive
lignin-based materials. Heating electrospun materials containing different ratios of the
fractions at different heating rates provided a means to control the morphology and properties
of lignin films. AFM also revealed that the differences in thermal flow behavior led to a
phase separated system reminiscent of SMMs. The structure of the lignin fractions was
shown to be different in terms of hydroxyl groups, interunit linkages, and molecular weight.
Our results suggest that Kraft lignin fractions may be employed to act as switch and rigid
phases in stimuli-responsive materials by taking advantage of the intrinsic differences in the
types and strengths of intermolecular interactions governing molecular mobility and the
supramolecular organization of lignin.
97
Chapter 6. Preparation and characterization of interconnected lignin-based
carbon nanofibre materials
6.1
Introduction:
In previous work we observed that thermostabilization of electrospun fibres
containing different amounts of F4 and F1-3 resulted in materials with different morphologies.
F4 had a very low thermal mobility and fibres containing only F4 maintained their shape after
thermostabilization. In contrast, F1-3 had a high thermal mobility and fibres containing F1-3
tended to soften and flow during thermostabilization, causing fibres to bond at their
intersections. It is hypothesized that a controlled degree of inter-fibre bonding might enable
enhancements in relevant material properties of electrospun lignin nonwoven fabrics
subjected to carbonization. In this study the effect of carbonization on the mechanical,
electrical, and surface properties of electrospun Kraft lignin-based fibres was investigated.
The properties of non-bonded (NBF) and bonded fibres (BF) were compared to determine if
inter-fibre bonding could be effective in enhancing the material properties of
thermostabilized and carbonized fabrics. Raman spectroscopy was also used to understand
the transformation occurring during carbonization treatment and also to compare the
carbonization processes of Kraft lignin and PAN in the temperature range 600-1000oC.
6.2
Carbonization of thermostabilized electrospun nonwovens:
The thermal softening of the fibres was previously observed in chapter 5 to depend on
the different thermal mobilities of the lignin fractions F4 and F1-3, the relative amount of each
fraction in the fibres, and the heating rate. In this study, thermostabilization at a temperature
of 250oC was carried out on electrospun fibres containing F4 to prepare NBF-250 materials
and thermostabilization of fibres containing a 70/30 wt/wt composition of F4/F1-3 was used to
prepare BF-250 materials. The same heating rate (5oC/min) was used during
thermostabilization for both materials. An SEM of the NBF-250 and BF-250 materials is
shown in Figure 6.1a and b, respectively.
98
Figure 6.1: (a) Non-bonded F4 fibres (NBF-250) and (b) Bonded 70/30 (w/w) F4/F1-3 fibres (BF-250) after
thermostabilization at 250oC. Scale bar = 20 m.
BF and NBF materials were subjected to further heat treatment under N2 at
temperatures (Tc) of 600, 800, or 1000oC in order to obtain carbonized materials. Figure 6.2
shows an SEM image of NBF and BF after carbonization to 1000oC. Similar appearances
were also observed for materials carbonized at 600 and 800oC (not shown). SEM confirmed
that the morphology obtained during thermostabilization was maintained after heating the
materials to 600, 800, or 1000oC and holding isothermally for 1 hour. The yield of
carbonization (based on the weight of thermostabilized fibres) was 48-52%, in agreement
with previous work on carbonization of Kraft lignin,98 and the yield values were identical for
carbonization at 600, 800, and 1000oC. No differences were observed in the yield when
comparing carbonization of the BF and NBF materials.
99
Figure 6.2: (a) Non-bonded lignin-based carbon fibres (NBF-1000) and (b) Bonded (w/w) F4/F1-3 carbon fibres
(BF-1000) after carbonization at 1000oC. Scale bar = 5 m.
SEM was used to measure the diameters of as-spun, thermostabilized, and carbonized
fibres, reported in Table 6.1. Since the BF materials were more difficult to measure
accurately in terms of fibre diameters due to slight alteration of the shapes due to inter-fibre
bonding, the values for only the NBF materials are reported. It should be noted that the asspun fibres had very similar diameters after electrospinning for the solution containing 28
wt% F4 compared to fibre electrospun from the solution containing 70/30 F4/F1-3 dissolved at
30 wt%. Also, the diameters of NBF-600, NBF-800, and NBF-1000 showed no significant
differences so the values for carbonized NBF materials can be considered representative of
all the different values of Tc. SEM showed that the fibre diameter decreased as a function of
heat treatment by a total of about 28% from 875 nm for the as-spun fibres to 634 nm for
carbonized fibres. Since it is known that smaller fibre diameters can be achieved through
electrospinning, goals of future studies should be to reduce the fibre diameter further and
evaluating any size effects in the mechanical properties. However, it should be noted that the
aim of the present study was to achieve good stability in the electrospinning process and
ensure that the fibres were free of beads and spray, and it was observed that there is a tradeoff between stability during electrospinning and fibre diameter. Parameters were selected to
achieve good quality fabrics to avoid complications in the interpretation of the mechanical
and electrical property characterization due to weak points or discontinuity in the fabrics that
could be caused by defects.
100
Table 6.1: Diameters of as-spun, thermostabilized, and carbonized lignin NBF materials.
Values are expressed as mean + one standard deviation. n = 200 fibres for each condition, 100 each for 2
different solutions
Sample
As-spun
Thermostabilized
Carbonized
6.3
Mean Fibre
Diameter (nm)
875 + 111
774 + 85
634 + 87
Effect of inter-fibre bonding on mechanical properties:
The mechanical properties of the NBF-250, BF-250, NBF-1000, and the carbonized
BF materials were measured in order to determine the effect of inter-fibre bonding on the
tensile properties before and after carbonization. The results of mechanical testing are
summarized in Table 6.2.
Table 6.2: Mechanical properties of bonded and non-bonded fabrics before and after carbonization at 1000 oC.
Values are expressed as mean + one standard deviation based on measurements on N samples.
Sample
NBF-1000
BF-1000
BF-800
BF-600
NBF-250
BF-250
Tensile
Strength
(MPa)
32.0 + 9.0
74.1 + 14.6
76.3 + 19.1
58.1 + 14.4
17.2 + 3.4
31.4 + 6.4
Elastic
modulus
(GPa)
4.8 + 0.6
4.1 + 1.4
2.4 + 0.5
2.0 + 0.5
0.8 + 0.06
1.3 + 0.09
% strain
at break
N
0.9 + 0.3
3.0 + 1.1
3.3 + 1.1
2.7 + 1.0
2.3 + 0.5
3.9 + 1.4
16
19
9
10
12
10
Mechanical testing showed that inter-fibre bonding significantly altered the
mechanical properties of the electrospun fabrics. NBF materials displayed rather low
mechanical performance, which was not unexpected due to the known fact that the
101
mechanical performance of lignin-based carbon fibres and precursors is considerably lower
than corresponding fibres derived from PAN.31 It is important to identify strategies to
improve the mechanical properties of lignin-based materials, and inter-fibre bonding was an
effective means to accomplish some improvements. BF materials had both a higher ultimate
tensile strength and higher elongation at break compared to NBF in the case of both
thermostabilized and carbonized fibres (Table 6.2). The elastic modulus values of the BF-250
materials were also higher than that for NBF-250 materials. Interestingly, this trend was
reversed after carbonization. NBF-1000 had a higher elastic modulus compared to BF-1000,
although there was considerable variability in the modulus values for the BF-1000 materials.
It should be mentioned that a difficulty in stabilization and carbonization of the materials was
their tendency to shrink, which sometimes caused the samples to slip out from between the
clamps holding them in place or break. This is an important point to consider in future
studies because maintaining tension without breaking the fibre during thermal treatment is
known to be critical for producing CF with good mechanical properties from PAN.110 Future
optimization studies to improve the dimensional stability of the materials during thermal
treatment are necessary. Nevertheless, the differences in BF vs. NBF materials were still
clear and noteworthy. It should also be noted that there was some difference between the
mechanical properties as a function of Tc (Table 6.2), but these differences were not as
pronounced as the differences observed between NBF and BF materials. A slight increase in
strength was observed comparing the properties of BF-600 to BF-800 and the strength
remained essentially unchanged comparing BF-800 to BF-1000.
Perhaps the most promising result in terms of mechanical properties was the clear
increase in the ductility of the BF materials as indicated by the higher strain at break. It is
well-known that brittleness is a shortcoming of lignin-based materials with high lignin
contents,325,326 and the flexibility of the thermostabilized BF materials was therefore
surprising. The fact that the ductility and flexibility of the BF materials was retained after
carbonization was also very interesting. Figure 6.3 shows photographs which demonstrate the
differences in flexibility between the carbonized NBF and BF materials. Fig. 6.3a shows the
result of applying only slight bending to the NBF-600 material, which resulted in breakage.
The BF-600 (Fig. 6.3b) was relatively more flexible, and the same trends were observed for
all values of Tc.
102
Figure 6.3: Photographs demonstrating the differences in flexibility between (a) NBF and (b) BF after
carbonization (Tc = 600oC). Slight bending resulted in breaking of the NBF material (a) while BF materials
were relatively flexible (b).
Tensile testing showed that BF-1000 materials could be stretched to strains averaging
3% before breaking, while the average strain at break value for BF-250 was 3.9%. In contrast
to the BF materials, the NBF-250 materials displayed an average strain at break of 2.3%,
while the NBF-1000 materials broke at strains below 1%. The results showed that
carbonization resulted in increased strength and modulus but decreased ductility for the NBF
and BF materials. However, inter-fibre bonding clearly increased ductility, especially in the
carbonized materials. For comparison, Kraft lignin-based carbon fibre single filaments were
reported to break around 1% strain,98 and thermally extruded Kraft lignin/PEO fibres
containing 25% or less PEO broke at 0.8% or lower strain, while 50/50 Kraft lignin/PEO
fibres could be elongated to 3.5% strain at the expense of tensile strength.192
Other researchers have also reported improvements in the mechanical properties of
electrospun fabrics through inter-fibre bonding.327,328 The mechanical property enhancement
in bonded fabrics may be due to improved load distribution among the fibre segments as
suggested for thermally point-bonded polypropylene nonwovens,329 though interestingly the
effect of bonding is known to vary depending on the strength of the fibres, and is strongly
dependent on bonding conditions.329,330 The different behaviors displayed by the lignin-based
103
BF and NBF materials may be due to different relative contributions of fibre stretching,
bending, and friction between sliding fibres, and the tendency of fibres to become oriented in
the stretching direction in accommodating imposed load.331 These factors are known to
influence the mechanical properties of electrospun nonwoven fabrics.332 While it is not
entirely clear what is the relationship between single fibre properties and the corresponding
NBF and BF materials, the differences in the BF and NBF are likely related to the single
fibre properties and how they respond to different types of loads. NBF materials consist of
much longer fibre segments and therefore experience a greater degree of sliding, bending,
and re-orientation. BF materials consisted of shorter interconnected segments that are not
capable of sliding with respect to one another and probably experience different degrees of
different types of stress compared to long fibre segments. The observed improvement in the
BF materials suggests that further optimization of thermal point bonding of electrospun
lignin fibres may be a promising route for further improvements in mechanical
performance.329,330 While the single fibre mechanical properties must be studied in more
detail to achieve further mechanical property improvements at the level of individual
filaments, it was clearly demonstrated that inter-fibre bonding enhanced the overall bulk
mechanical behavior of the electrospun fabrics.
6.4
Electrical conductivity of carbonized fabrics:
The electrical properties of CNFs are important for potential applications as
electrodes in batteries, solar cells, and supercapacitors. Inter-fibre bonding has also been
shown to lead to improvements in supercapacitor electrode performance using PAN-based
electrodes.120 The electrical conductivity () was therefore measured for BF and NBF
materials carbonized at Tc’s of 600, 800, and 1000oC and the values were compared to PANbased CNFs carbonized at 1000oC. The conductivity data is summarized in Table 6.3.
104
Table 6.3: Electrical conductivity  (S/cm) of carbonized samples. Values are expressed as mean + one
standard deviation. n = 10 for each condition
Sample
 (S/cm)
BF-1000
BF-800
BF-600
NBF-1000
NBF-800
NBF-600
PAN-1000
19.6 + 3.0
7.0 + 1.1
8.3 x 10-4 + 1.9 x10-4
2.3 + 0.4
0.6 + 0.1
1.2 x 10-4 + 4.7 x 10-5
9.6 + 1.7
NBF-600 and BF-600 materials had a very low  below 10-3 S/cm. In contrast, a
dramatic increase in  of more than 3 orders of magnitude was observed when the materials
were heated to 800 or 1000oC. The average  for NBF-800 and NBF-1000 were 0.6 and 2.3
S/cm, respectively, which was increased in the corresponding BF materials to 7 and 19.6
S/cm for BF-800 and BF-1000, respectively. The increase in  in the BF vs. NBF materials
may be due to the increased network connectivity in the BF materials allowing a greater
number of pathways for charge transport. For comparison, PAN-based nanofibres carbonized
at 1000oC were prepared and were observed to have an average  = 9.6 S/cm. SEM showed
that these PAN-based CNFs were not bonded at their intersections (image not shown), but it
has been reported that inter-fibre bonding increased  for bi-component
PAN/poly(vinylpyrrolidone)-based nanofibres.120 The  values of the lignin-based NBF-1000
and BF-1000 were also comparable to values reported for electrospun CNFs based on
phenolic resin = 5.96 S/cm)333 carbonized at 900oC, but lower than reported  values for
electrospun CNFs prepared from isotropic pitch ( = 63 S/cm for fibres carbonized at
1000oC).334 The results showed that  similar to that of other precursors can be achieved in
Kraft lignin-based CNFs and that inter-fibre bonding can increase the conductivity. The
electrical properties of lignin-based CNFs should therefore be studied in greater detail in
future work for possible application as electrodes.
105
6.5
BET surface area of carbonized fibres:
Since the surface area of carbon electrodes is also a very important parameter for
various applications, the surface area of the materials carbonized at different temperatures
was measured by the BET method. Generally, the SBET decreased slightly with increasing
cabonization temperature (Tc), but the values all fell in the range of 370-456 m2/g. The BF600, BF-800, and BF-1000 materials had SBET values of 456, 435, and 411 m2/g,
respectively, while the NBF-600, NBF-800, and NBF-1000 materials had SBET values of 450,
410, and 374 m2/g, respectively. These values are promising despite being slightly lower than
values reported for electrospun Alcell lignin-based CNFs243 and considerably lower than SBET
for porous carbon materials such as activated carbons, which can reach surface area values of
2000 m2/g.103 Given the slightly larger fibre diameters compared to previous reports242,243 and
the known fact that lignin is amenable to activation to increase its surface area,103,104,106–108
the surface areas reported here are a promising starting point for future studies.
6.6
Raman spectroscopy of carbonized fibres:
Raman spectroscopy is a powerful technique for the characterization of various forms
of disordered carbon.306,307,335–337 X-ray diffraction was initially used to characterize the
structure of electrospun fabrics after carbonization, but only a broad peak was observed (see
Appendix Figure A10), revealing little information regarding the effect of heat treatment on
fibre structure. Raman spectroscopy was instead used as a means to understand the structural
changes occurring during the carbonization process as a function of increasing Tc. Raman
spectra were recorded in the range 900-1800 cm-1 in order to compare the characteristic Dand G-bands typical of carbon materials for the different Tc values. The D-band corresponds
to the breathing modes of 6-carbon aromatic rings, while the G-band corresponds to in-plane
bond stretching motion of pairs of C sp2atoms in both rings and chains.306 A typical Raman
spectrum of carbonized lignin-based fibres is shown in Figure 6.4.
106
Figure 6.4: Typical Raman spectrum of carbonized lignin-based fibres in the wavenumber region 900-1800 cm1
. The D-band (~1310 cm-1) is fitted with a Lorentzian line shape and the G-band (1580 cm-1) is fitted with a
Breit-Wigner-Fano (BWF) lineshape, both shown in black, and the cumulative spectrum based on fitting is
shown in red.
In general the Raman spectra appeared fairly similar for the NBF and BF carbonized
at different Tc, consisting of a broad D-band around 1310 cm-1 partially overlapping with a
lower intensity G band around 1580 cm-1, typical of a disordered carbon material.306,307,335–337
Curve fitting with Lorentzian and Breit-Wigner-Fano (BWF) line shapes for the D- and Gbands, respectively, was used to extract information on peak heights to calculate the intensity
ratio (ID/IG), as well as width (FWHM), and position.306 Line plots of the ID/IG ratio, FWHM,
and band positions are shown in Figures 6.5, 6.6, and 6.7, respectively.
107
Figure 6.5: ID/IG from Raman spectra as a function of carbonization temperature for PAN, NBF, and BF. Error
bars represent plus/minus one standard deviation. PAN =
♦, NBF = ■, BF = ▲.
Figure 6.6: Full-width at half-maximum (FWHM, cm-1) of the (a) D-band and (b) G-band as a function of
♦
carbonization temperature from Raman spectra of PAN, NBF, and BF. PAN = , NBF =
bars represent plus/minus one standard deviation.
■, BF = ▲. Error
108
Figure 6.7: Positions (cm-1) of (a) D-band and (b) G-band as a function of carbonization temperature from
♦
■
Raman spectra of PAN, NBF, and BF. PAN = , NBF = , BF = ▲. Error bars represent plus/minus one
standard deviation.
At a given Tc, no significant differences in the ID/IG were observed between NBF and
BF, indicating that any differences in carbon structure due to the different lignin fractions is
indiscernible based solely on ID/IG (Fig. 6.5). Clear trends in the Raman spectra were
observed as a function of Tc. The ID/IG values for carbonized BF and NBF were identical and
clearly increased with increasing Tc, from 1.5 to 1.7 and 1.9 at Tc = 600, 800, and 1000oC,
respectively. This observation was in contrast with studies on PAN-derived fibres,122,298,338,339
where it reported that the ID/IG ratio decreases with increasing Tc in a similar temperature
range. Therefore, Raman spectra were also recorded for PAN nanofibres carbonized at 600,
800, and 1000oC and the spectra were processed in exactly the same manner as those for the
lignin-based nanofibres. The data for PAN is also shown graphically for ease of comparison
with lignin-based BF and NBF in Figure 6.5, 6.6, and 6.7.
The previously reported trend for ID/IG of PAN-based CNFs was also reproduced
when the spectra for PAN were processed in the same manner as those for the lignin-based
NBF and BF materials. PAN-based CNFs carbonized at 600oC had an ID/IG around 2.5, and
this value decreased with increasing Tc to 1.7 at 1000oC (Fig. 6.5). Furthermore, comparing
the trends in D and G band FWHM and peak positions also showed differences for PAN and
109
lignin-based CNFs. The FWHM of the D-band (Fig. 6.6a) increased from 600-800oC for
PAN and showed a slight but insignificant decrease from 800-1000oC. The D-band FWHM
decreased for the lignin-based BF and NBF as Tc increased. Interestingly, the D-band
FWHM started higher at 600oC for the BF material compared to that of the NBF material,
and the values converged to the same value after carbonization at 1000oC. The G-band
FWHM (Fig. 6.6b) also showed reversed trends as a function of Tc for PAN vs. lignin. The
G-band FWHM decreased with increasing Tc for PAN, but increased for BF and NBF. Also,
the G-band FWHM was higher in NBF vs. BF materials at all values of Tc.
In terms of peak position, increasing Tc resulted in a D-band shift to higher frequency
for PAN (Fig. 6.7a), whereas for NBF and BF, the D-band shifted to lower frequency
between 600-800oC and remained relatively constant from 800-1000oC. The G-band position
of PAN shifted to lower frequency from 600-800oC and shifted back to slightly higher
frequency from 800-1000oC (Fig. 6.7b). The G-band position for the BF and NBF also
displayed a curious difference whereby the BF G-band appeared to clearly shift to higher
frequency while the NBF G-band appeared to remain constant, although there was a
relatively high amount of variability in the G-band position of the NBF materials, indicated
by the relatively large error bars in Figure 6.7b.
Taken together, the results from Raman spectroscopy illustrate very clear differences
between the carbonization of PAN and Kraft lignin in the temperature range 600-1000oC.
There also appeared to be some differences between the NBF and BF materials. The threestage model developed by Ferrari and Robertson306,307 for describing Raman spectra of
amorphous and disordered carbon proved helpful in interpreting these differences. These
authors discussed a variety of published Raman data of disordered carbons in terms of an
“amorphization trajectory.” A key point from their publications was that the trends for the
ID/IG ratio for graphitic vs. amorphous carbons depend upon where along this amorphization
trajectory a particular carbon falls depending on its degree of graphitization or
amorphization. Based on the three-stage model, starting at a perfect graphite crystal, the
introduction of defects will increase the ID/IG ratio and initially shift the G-band to higher
frequency.306 However, upon transitioning from a graphite single crystal to a polycrystalline
material with very small nanographite crystals and further to a truly amorphous carbon, the
three-stage model predicts that this trend reverses and the ID/IG ratio will begin to decrease
110
with increasing amorphization due to a decrease in the number of ordered rings, while the Gband will shift to lower frequency.306 In the carbonization process it is reasonable to propose
that the carbon structures of lignin and PAN-based fibres essentially travel in reverse along
the amorphization trajectory as Tc increased and that lignin begins in a more amorphous state
relative to PAN, which exists as a semicrystalline material even before carbonization. PAN
may already exist as a nanocrystalline graphite at much lower carbonization temperature,
which explains why its ID/IG decreases to lower values. Lee et al. reported that the ID/IG ratio
of PAN fibres decreases as a function of increasing Tc from 400oC upward.339 The increase
of the ID/IG ratio observed for lignin-based fibres is therefore consistent with a different type
of transformation, from an amorphous carbon to a nanocrystalline graphite due to nucleation,
growth and clustering of aromatic rings (increased ordering of an amorphous carbon).306 The
dramatic increase in electrical conductivity of the materials comparing those processed at 600
and 800oC was also consistent with this type of transformation. The trends in ID/IG ratio
observed by Raman therefore identified a key difference in the carbonization behavior of
lignin and PAN.
The above hypothesis is also supported by the decrease in the FWHM of the D-band
for the lignin-based BF and NBF with increasing Tc (Fig. 6.6a). A decrease in the FWHM of
the D-band with increasing temperature was also reported for carbonization of wood,
cellulose, and organosolv lignin.340 Broadening of the D-band indicates a broader distribution
of clusters and rings other than 6-C aromatic rings.306 A decrease in D-band FWHM would
therefore be correlated with an increased number of 6-carbon aromatic rings in lignin-based
carbon as Tc increases from 600-1000oC. It is interesting that the PAN fibres showed the
opposite trend in the D-band FWHM between 600-800oC (Fig 6.6a) because that would
indicate that a broader distribution of rings with different numbers of C atoms forming for
PAN in this temperature range. The shifts in the D-band (Fig 6.7a) were also different for
lignin vs. PAN, indicating different processes are occurring during carbonization at the same
temperature.
The different trends in the G-band FWHM (Fig. 6.6b) and position (Fig 6.7b) are also
interesting but more difficult to interpret. The decreasing FWHM of the G-band for PAN
with increasing Tc seems to suggest a narrower distribution of sp2 hybridized carbon, but the
error associated with the determination of the G-band FWHM for PAN is large enough that
111
the G-band FWHM is arguably remaining fairly unchanged in the Tc range 600-1000oC. The
opposite trend in G-band FWHM was observed for lignin (Fig. 6.6b), and G-band FWHM
increased significantly by ~15 cm-1. The increase in G-band FWHM for lignin with
increasing may reflect an increase in the distribution of non-aromatic conjugated structures in
the lignin-based fibres, which may also be related to the observed increase in electrical
conductivity with increasing Tc. The three-stage model would also predict a shift to lower
frequency for the G-band of PAN and a shift to higher frequency for the G-band of lignin
based on the presumed location along the amorphization trajectory.306 G-band shifting to
higher frequency with increasing Tc would also be in agreement with clustering and sp2
carbon in chains. The predicted G-band shift to higher frequency was observed for BF but
curiously not for NBF, possibly indicating that the different structures of the two lignin
fractions incorporated into the BF materials have an effect on the formation of sp2 chains
and/or clusters. For PAN the shift to lower frequency was only observed between 600-800oC
but not from 800-1000oC. Future work using different laser energies and other
complementary techniques could be of use in elucidating these discrepancies. Nevertheless,
Raman spectroscopy proved to be extremely useful in interpreting and differentiating
between the structural changes occurring during carbonization of lignin and PAN-based
nanofibres, and should be considered a valuable tool for developing a better understanding of
the molecular level transformations occurring during the carbonization process.
6.7
Conclusion:
In this study, thermally induced inter-fibre bonding was found to be an effective
strategy for increasing the tensile strength, ductility, and electrical conductivity of Kraft
lignin-based carbon fibres obtained by electrospinning. While the mechanical properties were
still rather low compared to reported properties for PAN nanofibres, the ductility of the interbonded thermostabilized and carbonized materials was superior to single filament carbon
based on Kraft lignin. Further optimization of inter-fibre bonding conditions could
potentially allow further improvements in the mechanical properties of lignin-based
electrospun fibres. The electrical conductivity of the inter-bonded lignin-based fibres
carbonized at 1000oC was found to exceed that of non-bonded PAN-based carbon nanofibres
112
treated at the same carbonization temperature, suggesting that further study of the electrical
properties of lignin-based carbon nanofibre electrodes is warranted. The BET specific surface
area was slightly lower than previously reported values for Alcell lignin-based carbon
nanofibres, but can be considered a good starting point for future studies. Taken together the
material property characterizations suggest that lignin should be considered a candidate as a
precursor for flexible carbon electrode applications. A detailed analysis of Raman spectra
also shed new light on the differences in the carbonization behavior between lignin and PAN
in the temperature range of 600-1000oC. Interpretation of the trends in ID/IG ratio suggested
that lignin begins as a more amorphous carbon at 600oC and transforms to a nanocrystalline
graphite at temperatures of 800-1000oC, while PAN fibres have already formed a
nanocrystalline graphite structure at 600oC.
113
Chapter 7. Concluding Remarks
In these studies electrospinning of technical lignin was investigated for its potential to
spur the development of novel advanced lignin-based materials. Electrospinning has received
an enormous amount of attention in the research community in recent years, but
electrospinning of lignin, in contrast with synthetic polymers such as PAN, has been reported
by only a few research groups in peer-reviewed journals.242–245,341 The studies described in
the preceding chapters represent significant advancements in applying the fascinating
technique of electrospinning to processing of lignin. On the other hand, there is undoubtedly
room for further research on electrospinning of lignin with regard to each of the areas
discussed herein and new directions which have yet to be realized.
Chapter 3 began by presenting the finding that technical lignins dissolved in solution
have a strong tendency to resist the formation of uniform electrospun fibres and instead form
droplets. It is clear that extra steps must be taken in order to promote fibre formation during
electrospinning of lignin compared to synthetic polymers. The compact, branched structure
and the relatively low molecular weight of many technical lignins appears to hinder its ability
to form an entangled network of polymer chains in solution, which is known to facilitate
electrospinning of synthetic polymers. The studies presented in Chapters 3 suggest that a
relatively small amount of PEO added to the spinning solution is an effective strategy to
promote and control fibre formationduring electrospinning of technical lignins from
different sources.While thermal extrusion of softwood Kraft lignin was shown to be
relatively difficult compared to hardwood Kraft lignin in previous studies,98,294
electrospinning proved to be amenable to both types of lignin as well as other technical
lignins. This is an important point because preparation of lignin-based materials is
complicated by the fact that lignin is a naturally variable substance and bears modification
based on isolation processes as discussed in chapter 1.
Chapter 4 also showed that lignin can be purified in such a way as to allow it to form
fibres without the addition of synthetic polymers (section 4.5). This finding showed that
fractionation and purification of lignin strongly affects its processability by electrospinning.
The ability of the F4 fraction to form fibres may be related to its higher molecular weight and
114
strong intermolecular interactions governing the behaviour of this fraction, as studied in
Chapter 5. It was evident in the quality of the fibres obtained that fractionation strongly
improved the stability of the fibre formation process during electrospinning. These findings
suggest that heterogeneity in solubility and flow behaviours of oligomeric species of different
size and structure contributes to poor processing behaviour. Incorporating steps to obtain
more homogeneous lignin fractions during their isolation will provide new opportunities to
tune the processability of lignin.
The diameters achieved in these studies (~500-700 nm) are still somewhat large
compared to other electrospun fibres, which can reach diameters below 100 nm, even though
they are considerably smaller than thermally extruded lignin fibres. The diameters achieved
in these studies are probably still too large to observe mechanical property improvements
with decreasing diameter as has been reported for other materials.342,343,344 Smaller diameters
have been reported for lignin nanofibres, but the mechanical properties were not
characterized.242,243 The results of these studies are promising in that a tunable system was
developed to control the fibre diameter in the range > 500 nm, but it is critical to understand
the mechanism of instability leading to bead formation vs. fibre formation to optimize the
electrospinning process of lignin to achieve smaller fibre diameters, and to study the effect of
diameter on the internal fibre structure and material properties.
Another important question moving forward to future studies is: What is the most
practical way to implement electrospinning of lignin for real applications? Some well-known
limitations of the basic electrospinning apparatus for the production of commercially viable
products are the relatively low nanofibre production rate associated with using spinnerets,
inhomogeneous electric fields, and issues with spinneret clogging.345–348 Recent studies on
new variations of the basic electrospinning apparatus address these limitations by eliminating
the spinneret altogether. Instead of using spinnerets, “needleless” approaches for the
production of multiple electrospinning jets from the free surface of polymer solutions are
being explored.345–348 The first system for lignin electrospinning reported in published
literature was a coaxial spinneret approach employing a sheath fluid to stabilize the
electrospinning of Alcell lignin.242 Without the sheath fluid flowing at the proper flow rate,
Alcell lignin was observed to electrospray, as was also reported in chapter 3. A strong case
can be made that blending small amounts of synthetic polymer such as PEO or purifying
115
lignin so that it has the capacity to form fibres without the addition of other polymers is a
more practical strategy for lignin electrospinning compared to coaxial spinnerets. The
blending and purification studies reported herein are arguably much more likely to be
amenable to free-surface “needleless” electrospinning, and therefore could potentially allow
a higher rate of lignin nanofibre production. Whether or not free-surface electrospinning of
lignin can be implemented remains to be explored by future researchers, but should be
considered as an important area for future study.
The effect of the rheology of the spinning solution on the formation of electrospun
lignin fibres was also studied in detail with in chapter 4. This study represents a significant
step in generating an understanding of the elongational rheology of lignin and lignin/PEO
blends in electrospinning and other elongational flows. The results confirmed that polymer
blending is an effective strategy to modify the viscoelastic properties determining
processability of lignin. Understanding the elongational flow properties of lignin in solutions
and melt may prove helpful in guiding future attempts to modify the processability of lignin
into fibrous materials. Rheological studies showed that while adding a relatively small
amount of PEO to the spinning solution resulted in some minor change in the shear viscosity,
the effect of PEO addition on the elongational rheology of the spinning solution was
pronounced, even when only a small amount (less than 1%) of high molecular weight PEO
was added to the solution. CaBER provided a rapid means to characterize the elongational
fluid properties of lignin solutions, and a strong correlation between  and fibre diameter was
observed. CaBER should prove to be a valuable tool in the characterization of lignin and
lignin-polymer blend-based fluids for processing in electrospinning.
There are also many other possibilities for further exploration of elongational
rheology that were not addressed in these studies. Given that it is known that lignin and PEO
form a miscible polymer blend, the results suggest that specific intermolecular interactions
with small amounts of added polymers can be exploited to dramatically alter the processing
behaviour of lignin in elongational flow. However, a systematic study to explore the effect of
specific intermolecular interactions on elongational rheology was not performed in this work.
This represents an area which could be studied in more detail in future studies. It has also
been reported that the electrospinning jet itself can be used as an elongational rheometer for
studying elongational flow of polymer solutions under enormous strain rates which are
116
difficult or impossible to realize in other experiments.225,226 Electrospinning itself therefore
offers a unique opportunity to reveal new knowledge on the behavior of lignin and ligninpolymer blends undergoing strong stretching deformation, which could be beneficial both
from a theoretical and an applied perspective. For example, CaBER showed that lignin
solutions show Newtonian-like thinning behavior, but F4 was able to form fibres. It is
currently not known to what extent F4 solutions might become elastic under the strong
stretching characteristic of electrospinning. Using the jet itself as an elongational rheometer
could provide new insight into what happens to lignin macromolecules and whether they
behave in a Newtonian manner in a strongly stretched, electrified jet.
It was demonstrated in Chapter 5 that the versatility of electrospinning and the
lignin/PEO blend system allows the processing of combinations of lignin fractions with
different structure and properties into sub-micron diameter fibres. Fractionation of lignin by
solvent extraction and subsequent recombination of lignin fractions enabled the preparation
of a variety of interesting material morphologies including fibres, bonded nonwoven fibrous
networks, porous films, and smooth films which each might find interesting applications yet
to be explored. For example, a recent journal submission from G. Gao in our research group
demonstrated that thermostabilized electrospun fibres could be used as a substrate for
grafting new chemical functionalities onto lignin-based nanofibres. The findings in Chapter 5
underline the need to better understand and completely explore the possibilities of generating
lignin with desirable properties through fractionation. Much work remains to be done to
identify a scalable, inexpensive way to obtain fractionated lignin from pulping process
liquors, and understand how fractionation could help produce lignins more suited for
conversion to value-added products. Membrane filtration of lignin in black liquor can
apparently strongly influence the thermal processability of Kraft lignin320 in a similar way as
shown for solvent extraction, and it will be interesting to see to what extent membrane
filtration can be tuned and combined with innovative isolation strategies to generate pure,
homogeneous lignin fractions with special properties.
Chapter 5 also showed that using certain compositions of lignin fractions and proper
processing conditions, electrospun F4/F1-3 blend materials were shown to exist as novel
phase-separated systems which display reversible moisture-responsive shape change
capability. Fractionation of lignin was apparently a critical step in preparing these interesting
117
materials. Distinct properties of each of the two fractions as well as establishing an interface
between the two fractions were evidently key for achieving a reversible stimuliresponsiveness in electrospun materials. The potential applications for this new material
remain to be explored in future work, but a promising possibility is to design a humidity
sensor based on moisture-responsive lignin-based materials. In order to assess the viability of
this application, a more comprehensive study on the adsorption and desorption of water on
surface of these lignin-based SRM materials is needed to accurately describe the sensitivity
and kinetics of the moisture-responsive property and effect of cyclic adsorption and
desorption of moisture over the lifetime of the material.
In addition, the phase separation in moisture-responsive SRMs detected by AFM is in
itself a rather interesting phenomenon which raises new questions about the intrinsic
heterogeneity of lignin. D. Goring, a pioneer in the study of lignin’s chemistry and physical
properties, wrote in 1989 that the “properties of [lignin] macromolecules made soluble reflect
the properties of the network from which they are derived.”349 Recent findings that distinctly
different lignins are preferentially bound with different hemicelluloses167,168 show that lignin
may exist in distinctly different phases in wood. The finding presented in Chapter 5 that
lignin fractions extracted from a single commercially available Kraft lignin may act as
“switch” and “rigid” phases in an SRMs may indicate that the intrinsic heterogeneity of
lignin has a yet-to-be-explored function in the interesting mechanical properties of neverdried wood, and might be exploited further in the design of new lignin-based materials.254–256
The relative effects of the native structure vs. the changes resulting from isolation during
Kraft pulping remain to be elucidated to determine to what extent the moisture-responsive
property might be, as Goring suggested, a reflection of the supramacromolecular network
from which these lignin fractions are derived vs. a result of the changes imposed on lignin by
its isolation under the harsh conditions of Kraft delignification. While SRMs and SMMs are
highly active fields of research in advanced materials,58,81 lignin-based SRMs are a relatively
new.23 Incorporating other fractionation schemes, new processing strategies, and chemical
modification of lignin may open up new opportunities for design of advanced lignin-based
electrospun SRMs which can respond to a variety of different stimuli. The study described in
chapter 5 lays a foundation for the development of other types of advanced SRMs based on
phase-separated lignin systems, and potentially also for the development of true shape
118
memory materials which can be programmed into remembering two or more shapes. The
wide range of possibilities for design of new lignin-based SRMs and development of SRM
applications remain to be imagined and examined by future generations of researchers.
Chapter 6 showed that electrospinning is also an effective route to Kraft-lignin based
carbon nanofibres (CNFs). A somewhat disappointing result was that the mechanical
properties of these materials are likely unsuitable for reinforcement of structural composites,
but to put this finding in perspective, no lignin-based CFs have been reported to date which
can be considered suitable for structural applications.9,98,99 Future research must be conducted
to better understand the effect of the processing parameters and fibre diameter on the internal
fibre structure and single fibre properties. Nonetheless, the bonded nonwoven fibres prepared
in Chapter 5 were shown to maintain the BF morphology after carbonization at Tc up 1000oC
in Chapter 6, and inter-fibre bonding was shown to enhance both the mechanical and
electrical properties of lignin-based CNFs. These results suggest that further optimization of
fibre bonding conditions could further improve the properties of carbonized lignin-based BF
materials, given that it is known that the properties of bonded nonwoven materials are
sensitive to the bonding conditions, and the studies conducted in chapter 6 were not
optimized.329,330
The potential for preparation of electrospun, high surface area CNF electrodes for use
has been studied extensively.128 A combination of desirable surface and electrical properties
and suitable mechanical properties to prevent damage and deterioration of the electrodes are
critical to their performance, but will likely depend on the specific application. The BET
surface area (SBET) of the lignin-based CNFs reported in chapter 6 (~400 m2/g) are promising
given the fact that no activation process was used and the surface of the fibres appeared fairly
smooth under SEM (Fig. 6.2). Investigation of different physical and chemical activation
processes on the surface area, mechanical, and electrical properties is an important step to
advance these materials toward real applications. The study presented in Chapter 6 also
represents one of the first reports of the electrical conductivity of lignin-based CNFs
produced by electrospinning. Inter-fibre bonding of the lignin-based CNFs was shown to
enhance the bulk material’s electrical conductivity to values roughly twice the value for PAN
nanofibres, although they were compared to non-bonded PAN nanofibres. This result was
promising given the fact that PAN-based CNFs have been reported in numerous publications
119
for potential electrode applications, among others, which have been recently summarized in a
review by Inagaki and coworkers.128 Raman spectroscopy was also shown to be a powerful
technique for the comparison of the carbonization of lignin and PAN. Raman spectroscopy is
an essential tool for characterizing disorder in carbon materials,306,307,335 and future studies
using Raman to understand the evolution of order from disorder during thermal processing of
lignin could prove to be invaluable in understanding the structure-property relationships
governing the performance of lignin-based carbon materials. An interesting next step would
be to use Raman excitation at multiple wavelengths307 and also to explore the effect of
specific chemical characteristics on the carbonization process using model compounds or
different types of lignin.
Another essential next step is to evaluate the effect of the processing parameters for
lignin-based CNF preparation and the CNF material properties on the performance in specific
applications. The prospects for using the bonded CNF materials prepared in Chapter 6 are
very good. For comparison, Niu et al. recently investigated interconnected CNF networks
prepared by electrospinning combinations of PAN and polyvinylpyrrolidone for their
performance as supercapacitor electrodes and reported high specific capacitance of 221
F/g.120 Interestingly, they also reported that the highest capacitance was achieved with
materials having SBET of 531 m2/g, only slightly higher than the values reported in Chapter 6.
Interestingly, Niu and coworkers also found that materials with higher surface area had lower
capacitance, even though theory would suggest that the capacitance should be proportional to
the surface area accessible to ions.350 The slightly lower SBET values of the materials reported
in chapter 6 should not be considered a deterrent for investigating their potential as electrodes
in supercapacitors. Wang et al recently reported on capacitive deionization of NaCl solutions
using electrospun PAN-based CNF electrodes activated by simple CO2 activation (SBET =
712 m2/g) and demonstrated high electrosorption capacities of 4.64 mg Na+ per gram of
electrode material, which were higher than values reported for activated carbon and other
carbon materials.121 This exciting report suggests that high surface area carbon electrodes
produced by electrospinning could be useful in desalination of seawater. Potential
applications for interconnected lignin-based CNFs are therefore real and relevant and should
be investigated immediately.
120
The overriding question for utilization of lignin in value-added applications is
whether or not lignin-based materials can exhibit suitable, reliable performance that equals or
exceeds that of standard materials. In the case of mechanical performance, it is difficult to
envision lignin-based CNFs outperforming the mechanical properties of PAN. However,
electrospun lignin-based CNFs should absolutely be considered a promising replacement for
electrospun PAN-based CNFs in carbon electrode applications. A great deal of work remains
to be done, but it appears that further investigation of the electrospinning process for
development of lignin-based CNF materials is a promising avenue for the realization of
increased fundamental knowledge as well as commercial applications of lignin.
121
References
(1)
Ragauskas, A.J.; Williams, C.K.; Davison, B.H.; Britovsek, G.; Cairney, J.; Eckert,
C.A.; Frederick Jr. W.J.; Hallet, J.P.;Leak, D.J.; Liotta, C.L.; Mielenz, J.R.; Murphy,
R.; Templer, R.; Tschaplinski, T. The path forward for biofuels and biomaterials.
Science 2006, 311, 484–489.
(2)
Bozell, J. J.; Petersen, G. R. Technology development for the production of biobased
products from biorefinery carbohydrates—the US Department of Energy’s “Top 10”
revisited. Green Chemistry 2010, 12, 539–554.
(3)
Van Beilen, J. B.; Poirier, Y. Production of renewable polymers from crop plants. The
Plant journal : for cell and molecular biology 2008, 54, 684–701.
(4)
Effendi, A.; Gerhauser, H.; Bridgwater, A. V. Production of renewable phenolic resins
by thermochemical conversion of biomass: A review. Renewable and Sustainable
Energy Reviews 2008, 12, 2092–2116.
(5)
Stöcker, M. Biofuels and biomass-to-liquid fuels in the biorefinery: catalytic
conversion of lignocellulosic biomass using porous materials. Angewandte Chemie
(International ed. in English) 2008, 47, 9200–9211.
(6)
Ghatak, H. R. Biorefineries from the perspective of sustainability: Feedstocks,
products, and processes. Renewable and Sustainable Energy Reviews 2011, 15, 4042–
4052.
(7)
Zakzeski, J.; Bruijnincx, P. C. a; Jongerius, A. L.; Weckhuysen, B. M. The catalytic
valorization of lignin for the production of renewable chemicals. Chemical reviews
2010, 110, 3552–99.
(8)
Brodin, I.; Sjöholm, E.; Gellerstedt, G. Kraft lignin as feedstock for chemical
products: The effects of membrane filtration. Holzforschung 2009, 63, 290–297.
(9)
Gellerstedt, G.; Sjöholm, E.; Brodin, I. The Wood-Based Biorefinery : A Source of
Carbon Fiber? The Open Agriculture Journal 2010, 3, 119–124.
(10)
Lin, S.Y., Zhong, X. J. China Pulp Pap. 1990, 9, 45–53.
(11)
McCarthy, J.L.; Islam, A. Lignin chemistry, technology, and utilization: a brief
history. In Lignin: Historical, Biological, and Materials Perspectives.; Glasser, W.G.;
Northey, R.A.; Schultz, T. P., Ed.; ACS Symposium Series 742; American Chemical
Society: Washington, DC, 2000; pp. 2–99.
(12)
Gargulak, J.D., Lebo, S. E. Commercial use of lignin-based materials. In Lignin:
Historical, Biological, and Materials Perspectives.; Glasser, W.G., Northey, R.A.,
122
Schultz, T. P., Ed.; ACS Symposium Series 742; American Chemical Society:
Washington, DC, 2000; pp. 304–320.
(13)
Gosselink, R. J. a.; De Jong, E.; Guran, B.; Abächerli, a. Co-ordination network for
lignin—standardisation, production and applications adapted to market requirements
(EUROLIGNIN). Industrial Crops and Products 2004, 20, 121–129.
(14)
Li, Y.; Sarkanen, S. Thermoplastics with very high lignin contents. In Lignin:
Historical, Biological, and Materials Perspectives.; Glasser, W.G., Northey, R.A.,
Schultz, T. P., Ed.; ACS Symposium Series 742; American Chemical Society:
Washington, DC, 2000; pp. 351–366.
(15)
Ghosh, I.; Jain, R.K.; Glasser, W. G. Blends of biodegradable thermoplastics with
lignin esters. In Lignin: Historical, Biological, and Materials Perspectives.; Glasser,
W.G., Northey, R.A, Schultz, T. P., Ed.; ACS Symposium Series 742; American
Chemical Society: Washington, DC, 2000; pp. 331–350.
(16)
Ono, H.-K.; Sudo, K. Wood adhesives from lignin: a way to use lignin from steamexplosion process. In Lignin: Properties and Materials; Glasser, W.G., Sarkanen, S.,
Ed.; ACS Symposium Series 397, American Chemical Society: Washington, DC,
1989; pp. 334–345.
(17)
Lindberg, J.J.; Kuusela, T.A.; Levon, K. Specialty polymers from lignin. In Lignin:
Properties and Materials; Glasser, W.G., Sarkanen, S., Ed.; ACS Symposium Series
397, American Chemical Society: Washington, DC, 1989; pp. 190–204.
(18)
Nemoto, T.; Konishi, G.; Tojo, Y.; An, Y. C.; Funaoka, M. Functionalization of
Lignin : Synthesis of Lignophenol- graft-Poly ( 2-ethyl-2-oxazoline ) and Its
Application to Polymer Blends with Commodity Polymers. Journal of Applied
Polymer Science 2012, 123, 2636–2642.
(19)
Thanh Binh, N. T.; Luong, N. D.; Kim, D. O.; Lee, S. H.; Kim, B. J.; Lee, Y. S.; Nam,
J.-D. Synthesis of Lignin-Based Thermoplastic Copolyester Using Kraft Lignin as a
Macromonomer. Composite Interfaces 2009, 16, 923–935.
(20)
Li, Y.; Ragauskas, A. J. Kraft Lignin-Based Rigid Polyurethane Foam. Journal of
Wood Chemistry and Technology 2012, 32, 210–224.
(21)
Chung, H.; Washburn, N. R. Improved Lignin Polyurethane Properties with Lewis
Acid Treatment. ACS applied materials & interfaces 2012.
(22)
Hatakeyama, H.; Hatakeyama, T. Lignin Structure , Properties , and Applications.
Advances in Polymer Science 2010, 232, 1–63.
(23)
Kim, Y. S.; Kadla, J. F. Preparation of a thermoresponsive lignin-based biomaterial
through atom transfer radical polymerization. Biomacromolecules 2010, 11, 981–8.
123
(24)
Stanzione III, J. F.; Sadler, J. M.; La Scala, J. J.; Reno, K. H.; Wool, R. P. Vanillinbased resin for use in composite applications. Green Chemistry 2012, 14, 2346.
(25)
Pérez, J. M.; Oliet, M.; Alonso, M. V.; Rodríguez, F. Cure kinetics of lignin–novolac
resins studied by isoconversional methods. Thermochimica Acta 2009, 487, 39–42.
(26)
Guigo, N.; Mija, A.; Vincent, L.; Sbirrazzuoli, N. Eco-friendly composite resins based
on renewable biomass resources: Polyfurfuryl alcohol/lignin thermosets. European
Polymer Journal 2010, 46, 1016–1023.
(27)
Passauer, L.; Fischer, K.; Liebner, F. Preparation and physical characterization of
strongly swellable oligo (oxyethylene) lignin hydrogels. Holzforschung 2011, 65, 309–
317.
(28)
Rutledge, G. C.; Fridrikh, S. V Formation of fibers by electrospinning. Advanced drug
delivery reviews 2007, 59, 1384–91.
(29)
Reneker, D. H.; Yarin, A. L. Electrospinning jets and polymer nanofibers. Polymer
2008, 49, 2387–2425.
(30)
Greiner, A.; Wendorff, J. H. Electrospinning: a fascinating method for the preparation
of ultrathin fibers. Angewandte Chemie (International ed. in English) 2007, 46, 5670–
703.
(31)
Arshad, S. N.; Naraghi, M.; Chasiotis, I. Strong carbon nanofibers from electrospun
polyacrylonitrile. Carbon 2011, 49, 1710–1719.
(32)
Liu, J.; Yue, Z.; Fong, H. Continuous nanoscale carbon fibers with superior
mechanical strength. Small 2009, 5, 536–42.
(33)
Chen, L.; Bromberg, L.; Hatton, T. A.; Rutledge, G. C. Electrospun cellulose acetate
fibers containing chlorhexidine as a bactericide. Polymer 2008, 49, 1266–1275.
(34)
Aluigi, A.; Vineis, C.; Varesano, A.; Mazzuchetti, G.; Ferrero, F.; Tonin, C. Structure
and properties of keratin/PEO blend nanofibres. European Polymer Journal 2008, 44,
2465–2475.
(35)
Varesano, A.; Aluigi, A.; Vineis, C.; Tonin, C. Study on the Shear Viscosity Behavior
of Keratin / PEO Blends for Nanofibre Electrospinning. Journal of Polymer Science
Part B: Polymer Physics 2008, 46, 1193–1201.
(36)
Nie, H.; He, A.; Wu, W.; Zheng, J.; Xu, S.; Li, J.; Han, C. C. Effect of poly(ethylene
oxide) with different molecular weights on the electrospinnability of sodium alginate.
Polymer 2009, 50, 4926–4934.
124
(37)
Klossner, R. R.; Queen, H. a; Coughlin, A. J.; Krause, W. E. Correlation of chitosan’s
rheological properties and its ability to electrospin. Biomacromolecules 2008, 9, 2947–
53.
(38)
Ayutsede, J.; Gandhi, M.; Sukigara, S.; Micklus, M.; Chen, H.-E.; Ko, F. Regeneration
of Bombyx mori silk by electrospinning. Part 3: characterization of electrospun
nonwoven mat. Polymer 2005, 46, 1625–1634.
(39)
Ayutsede, J.; Gandhi, M.; Sukigara, S.; Ye, H.; Hsu, C.; Gogotsi, Y.; Ko, F. Carbon
nanotube reinforced Bombyx mori silk nanofibers by the electrospinning process.
Biomacromolecules 2006, 7, 208–14.
(40)
Gupta, P.; Elkins, C.; Long, T. E.; Wilkes, G. L. Electrospinning of linear
homopolymers of poly(methyl methacrylate): exploring relationships between fiber
formation, viscosity, molecular weight and concentration in a good solvent. Polymer
2005, 46, 4799–4810.
(41)
Inai, R.; Kotaki, M.; Ramakrishna, S. Structure and properties of electrospun PLLA
single nanofibres. Nanotechnology 2005, 16, 208–13.
(42)
Ramaseshan, R.; Sundarrajan, S.; Jose, R.; Ramakrishna, S. Nanostructured ceramics
by electrospinning. Journal of Applied Physics 2007, 102, 111101.
(43)
Gandhi, M.; Yang, H.; Shor, L.; Ko, F. Post-spinning modification of electrospun
nanofiber nanocomposite from Bombyx mori silk and carbon nanotubes. Polymer
2009, 50, 1918–1924.
(44)
Andrew, J. S.; Mack, J. J.; Clarke, D. R. Electrospinning of polyvinylidene difluoridebased nanocomposite fibers. Journal of Materials Research 2011, 23, 105–114.
(45)
Chen, D.; Liu, T.; Zhou, X.; Tjiu, W. C.; Hou, H. Electrospinning fabrication of high
strength and toughness polyimide nanofiber membranes containing multiwalled
carbon nanotubes. The journal of physical chemistry. B 2009, 113, 9741–8.
(46)
Huang, Z.-M.; Zhang, Y.-Z.; Kotaki, M.; Ramakrishna, S. A review on polymer
nanofibers by electrospinning and their applications in nanocomposites. Composites
Science and Technology 2003, 63, 2223–2253.
(47)
Im, J. S.; Park, S.-J.; Kim, T.; Lee, Y.-S. Hydrogen storage evaluation based on
investigations of the catalytic properties of metal/metal oxides in electrospun carbon
fibers. International Journal of Hydrogen Energy 2009, 34, 3382–3388.
(48)
Demir, M. M.; Gulgun, M. A.; Menceloglu, Y. Z.; Abramchuk, S. S.; Makhaeva, E.
E.; Khokhlov, A. R.; Matveeva, V. G.; Sulman, M. G. Palladium Nanoparticles by
Electrospinning from Poly ( acrylonitrile-co-acrylic acid ) -PdCl 2 Solutions .
125
Relations between Preparation Conditions , Particle Size , and Catalytic Activity.
Macromolecules 2004, 37, 1787–1792.
(49)
Agarwal, S.; Wendorff, J. H.; Greiner, A. Use of electrospinning technique for
biomedical applications. Polymer 2008, 49, 5603–5621.
(50)
Prabhakaran, M. P.; Venugopal, J. R.; Ramakrishna, S. Mesenchymal stem cell
differentiation to neuronal cells on electrospun nanofibrous substrates for nerve tissue
engineering. Biomaterials 2009, 30, 4996–5003.
(51)
Venugopal, J.; Vadgama, P.; Kumar, T. S. S.; Ramakrishna, S. Biocomposite
nanofibres and osteoblasts for bone tissue engineering. Nanotechnology 2007, 18,
055101.
(52)
Sharma, S.; Sharma, A.; Cho, Y.-K.; Madou, M. Increased graphitization in
electrospun single suspended carbon nanowires integrated with carbon-MEMS and
carbon-NEMS platforms. ACS applied materials & interfaces 2012, 4, 34–9.
(53)
Norris, I. D.; Shaker, M. M.; Ko, F. K.; Macdiarmid, A. G. Electrostatic fabrication of
ultrafine conducting fibers : polyanilinerpolyethylene oxide blends. Synthetic Metals
2000, 114, 109–114.
(54)
Ahn, S.; Kasi, R. M.; Kim, S.-C.; Sharma, N.; Zhou, Y. Stimuli-responsive polymer
gels. Soft Matter 2008, 4, 1151.
(55)
Ionov, L. Actively-moving materials based on stimuli-responsive polymers. Journal of
Materials Chemistry 2010, 20, 3382.
(56)
Klinger, D.; Landfester, K. Stimuli-Responsive Microgels for the Loading and Release
of Functional Compounds: Fundamental Concepts and Applications. Polymer 2012,
Article in, 1–23.
(57)
Seiffert, S.; Sprakel, J. Physical chemistry of supramolecular polymer networks.
Chemical Society reviews 2012, 41, 909–30.
(58)
Stuart, M. a C.; Huck, W. T. S.; Genzer, J.; Müller, M.; Ober, C.; Stamm, M.;
Sukhorukov, G. B.; Szleifer, I.; Tsukruk, V. V; Urban, M.; Winnik, F.; Zauscher, S.;
Luzinov, I.; Minko, S. Emerging applications of stimuli-responsive polymer materials.
Nature materials 2010, 9, 101–13.
(59)
Such, G. K.; Johnston, A. P. R.; Caruso, F. Engineered hydrogen-bonded polymer
multilayers: from assembly to biomedical applications. Chemical Society reviews
2011, 40, 19–29.
(60)
Yan, X.; Wang, F.; Zheng, B.; Huang, F. Stimuli-responsive supramolecular
polymeric materials. Chemical Society reviews 2012, 0012.
126
(61)
Lendlein, A.; Kelch, S. Shape-Memory Polymers. Angewandte Chemie (International
ed. in English) 2002, 41, 2034–2057.
(62)
Lendlein, A.; Jiang, H.; Junger, O.; Langer, R. Light-induced shape-memory polymers
¨. Nature 2005, 434, 879–882.
(63)
Leng, J.; Lv, H.; Liu, Y.; Du, S. Synergic effect of carbon black and short carbon fiber
on shape memory polymer actuation by electricity. Journal of Applied Physics 2008,
104, 104917.
(64)
Mohr, R.; Kratz, K.; Weigel, T.; Lucka-Gabor, M.; Moneke, M.; Lendlein, a Initiation
of shape-memory effect by inductive heating of magnetic nanoparticles in
thermoplastic polymers. Proceedings of the National Academy of Sciences of the
United States of America 2006, 103, 3540–5.
(65)
Chen, S.; Hu, J.; Yuen, C.; Chan, L. Novel moisture-sensitive shape memory
polyurethanes containing pyridine moieties. Polymer 2009, 50, 4424–4428.
(66)
Du, H.; Zhang, J. Shape memory polymer based on chemically cross-linked poly(vinyl
alcohol) containing a small number of water molecules. Colloid and Polymer Science
2010, 288, 15–24.
(67)
Du, H.; Zhang, J. Solvent induced shape recovery of shape memory polymer based on
chemically cross-linked poly(vinyl alcohol). Soft Matter 2010, 6, 3370–3376.
(68)
Fan, K.; Huang, W.M.; Wang, C.C.; Ding, Z.; Zhao, Y.; Purnawali, H.; Liew, K.C.
Zheng, L. X. Water-responsive shape memory hybrid: Design concept and
demonstration. Express Polymer Letters 2011, 5, 409–416.
(69)
Huang, W. M.; Yang, B.; An, L.; Li, C.; Chan, Y. S. Water-driven programmable
polyurethane shape memory polymer: Demonstration and mechanism. Applied Physics
Letters 2005, 86, 114105.
(70)
Huang, W. M.; Yang, B.; Zhao, Y.; Ding, Z. Thermo-moisture responsive
polyurethane shape-memory polymer and composites: a review. Journal of Materials
Chemistry 2010, 20, 3367–3381.
(71)
Chae Jung, Y.; Hwa So, H.; Whan Cho, J. Water Responsive Shape Memory
Polyurethane Block Copolymer Modified with Polyhedral Oligomeric Silsesquioxane.
Journal of Macromolecular Science, Part B: Physics 2006, 45, 453–461.
(72)
Mendez, J.; Annamalai, P. K.; Eichhorn, S. J.; Rusli, R.; Rowan, S. J.; Foster, E. J.;
Weder, C. Bioinspired Mechanically Adaptive Polymer Nanocomposites with WaterActivated Shape-Memory Effect. Macromolecules 2011, 44, 6827–6835.
127
(73)
Yang, B.; Min Huang, W.; Li, C.; Hoe Chor, J. Effects of moisture on the glass
transition temperature of polyurethane shape memory polymer filled with nano-carbon
powder. European Polymer Journal 2005, 41, 1123–1128.
(74)
Yang, B.; Huang, W. M.; Li, C.; Li, L. Effects of moisture on the thermomechanical
properties of a polyurethane shape memory polymer. Polymer 2006, 47, 1348–1356.
(75)
Yang, B.; Huang, W. M.; Li, C.; Lee, C. M.; Li, L. On the effects of moisture in a
polyurethane shape memory polymer. Smart Materials and Structures 2004, 13, 191–
195.
(76)
Zhu, Y.; Hu, J.; Luo, H.; Young, R. J.; Deng, L.; Zhang, S.; Fan, Y.; Ye, G. Rapidly
switchable water-sensitive shape-memory cellulose/elastomer nano-composites. Soft
Matter 2012, 8, 2509–2517.
(77)
Lendlein, A.; Langer, R. Biodegradable, elastic shape-memory polymers for potential
biomedical applications. Science (New York, N.Y.) 2002, 296, 1673–6.
(78)
Kunzelman, J.; Chung, T.; Mather, P. T.; Weder, C. Shape memory polymers with
built-in threshold temperature sensors. Journal of Materials Chemistry 2008, 18, 1082.
(79)
Lu, H.; Gou, J.; Leng, J.; Du, S. Magnetically aligned carbon nanotube in nanopaper
enabled shape-memory nanocomposite for high speed electrical actuation. Applied
Physics Letters 2011, 98, 174105.
(80)
Liu, C.; Qin, H.; Mather, P. T. Review of progress in shape-memory polymers.
Journal of Materials Chemistry 2007, 17, 1543.
(81)
Leng, J.; Lan, X.; Liu, Y.; Du, S. Shape-memory polymers and their composites:
Stimulus methods and applications. Progress in Materials Science 2011, 56, 1077–
1135.
(82)
Ratna, D.; Karger-Kocsis, J. Recent advances in shape memory polymers and
composites: a review. Journal of Materials Science 2008, 43, 254–269.
(83)
Pandini, S.; Passera, S.; Messori, M.; Paderni, K.; Toselli, M.; Gianoncelli, a.;
Bontempi, E.; Riccò, T. Two-way reversible shape memory behaviour of crosslinked
poly(ε-caprolactone). Polymer 2012, 53, 1915–1924.
(84)
Jeong, H.M.; Ahn, B.K.; Kim, B. K. Miscibility and shape memory e € ect of
thermoplastic polyurethane blends with phenoxy resin. European Polymer Journal
2001, 37, 2245–2252.
(85)
Zhang, H.; Wang, H.; Zhong, W.; Du, Q. A novel type of shape memory polymer
blend and the shape memory mechanism. Polymer 2009, 50, 1596–1601.
128
(86)
Kurahashi, E.; Sugimoto, H.; Nakanishi, E.; Nagata, K.; Inomata, K. Shape memory
properties of polyurethane/poly(oxyethylene) blends. Soft Matter 2012, 8, 496–503.
(87)
Jeong, H .M.; Song, J.H.; Lee, S. Y. .; Kim, B. K. Miscibility and shape memory
property of poly (vinyl chloride )/ thermoplastic polyurethane blends. Journal of
Materials Science 2001, 36, 5457–5463.
(88)
Luo, X.; Mather, P. T. Triple-Shape Polymeric Composites (TSPCs). Advanced
Functional Materials 2010, 20, 2649–2656.
(89)
D’hollander, S.; Van Assche, G.; Van Mele, B.; Du Prez, F. Novel synthetic strategy
toward shape memory polyurethanes with a well-defined switching temperature.
Polymer 2009, 50, 4447–4454.
(90)
Bellin, I.; Kelch, S.; Langer, R.; Lendlein, a Polymeric triple-shape materials.
Proceedings of the National Academy of Sciences of the United States of America
2006, 103, 18043–7.
(91)
Ware, T.; Hearon, K.; Lonnecker, A.; Wooley, K. L.; Maitland, D. J.; Voit, W. TripleShape Memory Polymers Based on Self-Complementary Hydrogen Bonding.
Macromolecules 2012, 45, 1062–1069.
(92)
Ahn, S.; Kasi, R. M. Exploiting Microphase-Separated Morphologies of Side-Chain
Liquid Crystalline Polymer Networks for Triple Shape Memory Properties. Advanced
Functional Materials 2011, 21, 4543–4549.
(93)
Auad, L.; Contos, V. S.; Nutt, S.; Aranguren, M. I.; Marcovich, N. E. Characterization
of nanocellulose- reinforced shape memory polyurethanes. 2008, 659, 651–659.
(94)
Aoki, D.; Teramoto, Y.; Nishio, Y. SH-containing cellulose acetate derivatives:
preparation and characterization as a shape memory-recovery material.
Biomacromolecules 2007, 8, 3749–57.
(95)
Skrzeszewska, P. J.; Jong, L. N.; De Wolf, F. a; Cohen Stuart, M. a; Van der Gucht, J.
Shape-memory effects in biopolymer networks with collagen-like transient nodes.
Biomacromolecules 2011, 12, 2285–92.
(96)
Huang, C.; Soenen, S. J.; Rejman, J.; Lucas, B.; Braeckmans, K.; Demeester, J.; De
Smedt, S. C. Stimuli-responsive electrospun fibers and their applications. Chemical
Society reviews 2011, 40, 2417–34.
(97)
Kubo, S.; Kadla, J. F. Carbon fibers from lignin-recyclable plastic blends. In
Encyclopedia of Chemical Processing; Lee, S., Ed.; Taylor and Francis Group: New
York, 2006; Vol. 1, pp. 317–331.
129
(98)
Kadla, J.F.; Kubo, S.;Venditti, R.A.; Gilbert, R.D.; Compere, A.L.; Griffith, W.
Lignin-based carbon fibers for composite fiber applications. Carbon 2002, 40, 2913–
2920.
(99)
Baker, D. A.; Gallego, N. C.; Baker, F. S. On the Characterization and Spinning of an
Organic- Purified Lignin Toward the Manufacture of Low-Cost Carbon Fiber. Journal
of Applied Polymer Science 2012, 124, 227–234.
(100) Kubo, S.; Kadla, J. F. Lignin-based Carbon Fibers: Effect of Synthetic Polymer
Blending on Fiber Properties. Journal of Polymers and the Environment 2005, 13, 97–
105.
(101) Uraki, Y.; Nakatani, A.; Kubo, S.; Sano, Y. Preparation of activated carbon fibers with
large specific surface area from softwood acetic acid lignin. Journal of Wood Science
2001, 47, 465–469.
(102) Kubo, S.; Yoshida, T.; Kadla, J. F. Surface Porosity of Lignin/PP Blend Carbon
Fibers. Journal of Wood Chemistry and Technology 2007, 27, 257–271.
(103) Hayashi, J.; Kazehaya, A.; Muroyama, K.; Watkinson, A. P. Preparation of activated
carbon from lignin by chemical activation. Carbon 2000, 38, 1873–1878.
(104) Kijima, M.; Hirukawa, T.; Hanawa, F.; Hata, T. Thermal conversion of alkaline lignin
and its structured derivatives to porous carbonized materials. Bioresource technology
2011, 102, 6279–85.
(105) Shen, Q.; Zhang, T.; Zhang, W.; Chen, S.; Mezgebe, M. Lignin-Based Activated
Carbon Fibers and Controllable Pore Size and Properties. Journal of Applied Polymer
Science 2011, 121, 989–994.
(106) Babeł, K.; Jurewicz, K. KOH activated lignin based nanostructured carbon exhibiting
high hydrogen electrosorption. Carbon 2008, 46, 1948–1956.
(107) Carrott, P. J. M.; Carrott, M. M. L. R. Using alkali metals to control reactivity and
porosity during physical activation of demineralised kraft lignin. Carbon 2009, 47,
1012–1017.
(108) Suhas; Carrott, P. J. M.; Ribeiro Carrott, M. M. L. Lignin--from natural adsorbent to
activated carbon: a review. Bioresource technology 2007, 98, 2301–12.
(109) Morgan, P. Carbon fibers and their composites; Taylor and Francis: Boca Raton,
2005.
(110) Frank, E.; Hermanutz, F.; Buchmeiser, M. R. Carbon Fibers: Precursors,
Manufacturing, and Properties. Macromolecular Materials and Engineering 2012,
297, 493–501.
130
(111) Osmond, N. M. Activated Carbon Fibre Adsorbent Materials. Adsorption Science &
Technology 2000, 18, 529–539.
(112) Calvino-Casilda, V.; López-Peinado, a. J.; Durán-Valle, C. J.; Martín-Aranda, R. M.
Last Decade of Research on Activated Carbons as Catalytic Support in Chemical
Processes. Catalysis Reviews 2010, 52, 325–380.
(113) Yang, Y.; Chiang, K.; Burke, N. Porous carbon-supported catalysts for energy and
environmental applications: A short review. Catalysis Today 2011, 178, 197–205.
(114) Saito, N.; Aoki, K.; Usui, Y.; Shimizu, M.; Hara, K.; Narita, N.; Ogihara, N.;
Nakamura, K.; Ishigaki, N.; Kato, H.; Haniu, H.; Taruta, S.; Kim, Y. A.; Endo, M.
Application of carbon fibers to biomaterials: a new era of nano-level control of carbon
fibers after 30-years of development. Chemical Society reviews 2011, 40, 3824–34.
(115) Chung, D. D. L. Review Electrical applications of carbon materials. Journal of
Materials Science 2004, 39, 2645–2661.
(116) Pollak, E.; Levy, N.; Eliad, L.; Salitra, G.; Soffer, A.; Aurbach, D. Review on
Engineering and Characterization of Activated Carbon Electrodes for Electrochemical
Double Layer Capacitors and Separation Processes. Israel Journal of Chemistry 2008,
48, 287–303.
(117) Snyder, J. F.; Wong, E. L.; Hubbard, C. W. Evaluation of Commercially Available
Carbon Fibers, Fabrics, and Papers for Potential Use in Multifunctional Energy
Storage Applications. Journal of The Electrochemical Society 2009, 156, A215.
(118) Huang, Z.-H.; Wang, M.; Wang, L.; Kang, F. Relation between the charge efficiency
of activated carbon fiber and its desalination performance. Langmuir : the ACS journal
of surfaces and colloids 2012, 28, 5079–84.
(119) Kim, B.-H.; Yang, K. S.; Ferraris, J. P. Highly conductive, mesoporous carbon
nanofiber web as electrode material for high-performance supercapacitors.
Electrochimica Acta 2012, 75, 325–331.
(120) Niu, H.; Zhang, J.; Xie, Z.; Wang, X.; Lin, T. Preparation, structure and
supercapacitance of bonded carbon nanofiber electrode materials. Carbon 2011, 49,
2380–2388.
(121) Wang, G.; Pan, C.; Wang, L.; Dong, Q.; Yu, C.; Zhao, Z.; Qiu, J. Activated carbon
nanofiber webs made by electrospinning for capacitive deionization. Electrochimica
Acta 2012, 69, 65–70.
(122) Kim, C.; Yang, K. S.; Kojima, M.; Yoshida, K.; Kim, Y. J.; Kim, Y. a.; Endo, M.
Fabrication of Electrospinning-Derived Carbon Nanofiber Webs for the Anode
131
Material of Lithium-Ion Secondary Batteries. Advanced Functional Materials 2006,
16, 2393–2397.
(123) Lee, B.-S.; Son, S.-B.; Park, K.-M.; Yu, W.-R.; Oh, K.-H.; Lee, S.-H. Anodic
properties of hollow carbon nanofibers for Li-ion battery. Journal of Power Sources
2012, 199, 53–60.
(124) Dimesso, L.; Spanheimer, C.; Jaegermann, W.; Zhang, Y.; Yarin, a. L. LiFePO4 – 3D
carbon nanofiber composites as cathode materials for Li-ions batteries. Journal of
Applied Physics 2012, 111, 064307.
(125) Chen, S.; Hou, H.; Harnisch, F.; Patil, S. a.; Carmona-Martinez, A. a.; Agarwal, S.;
Zhang, Y.; Sinha-Ray, S.; Yarin, A. L.; Greiner, A.; Schröder, U. Electrospun and
solution blown three-dimensional carbon fiber nonwovens for application as
electrodes in microbial fuel cells. Energy & Environmental Science 2011, 4, 1417.
(126) Li, M.; Zhao, S.; Han, G.; Yang, B. Electrospinning-derived carbon fibrous mats
improving the performance of commercial Pt/C for methanol oxidation. Journal of
Power Sources 2009, 191, 351–356.
(127) Joshi, P.; Zhang, L.; Chen, Q.; Galipeau, D.; Fong, H.; Qiao, Q. Electrospun carbon
nanofibers as low-cost counter electrode for dye-sensitized solar cells. ACS applied
materials & interfaces 2010, 2, 3572–7.
(128) Inagaki, M.; Yang, Y.; Kang, F. Carbon nanofibers prepared via electrospinning.
Advanced materials (Deerfield Beach, Fla.) 2012, 24, 2547–66.
(129) Tagawa, T.; Miyata, T. Size effect on tensile strength of carbon fibers. 1997, 238,
336–342.
(130) Tanaka, T.; Nakayama, H.; Sakaida, A.; Horikawa, N. Estimation of tensile strength
distribution for carbon fiber with diameter variation along fiber. Materials Science
Research International 1999, 5, 90–97.
(131) Lignins: Occurrence, formation, structure, and reactions.; Sarkanen, K.V., Ludwig, C.
H., Ed.; Wiley Interscience: New York, 1971.
(132) Boerjan, W.; Ralph, J.; Baucher, M. Lignin biosynthesis. Annual review of plant
biology 2003, 54, 519–46.
(133) Ralph, J.; Lundquist, K.; Brunow, G.; Lu, F.; Kim, H.; Schatz, P. F.; Marita, J. M.;
Hatfield, R. D.; Ralph, S. a.; Christensen, J. H.; Boerjan, W. Lignins: Natural
polymers from oxidative coupling of 4-hydroxyphenyl- propanoids. Phytochemistry
Reviews 2004, 3, 29–60.
(134) Ralph, J. Hydroxycinnamates in lignification. Phytochemistry Reviews 2009, 9, 65–83.
132
(135) Brown, S.A.; Neish, A. C. Shikimic acid as a precursor in lignin biosynthesis. Nature
1955, 175, 688.
(136) Sarkanen, K. V. Precursors and their polymerization. In Lignins: Occurrence,
formation, structure, and reactions; Sarkanen, K.V., Ludwig, C. H., Ed.; Wiley
Interscience: New York, 1971; pp. 95–163.
(137) Humphreys, J. M.; Chapple, C. Rewriting the lignin roadmap. Current opinion in plant
biology 2002, 5, 224–9.
(138) Li, L.; Popko, J. L.; Umezawa, T.; Chiang, V. L. 5-Hydroxyconiferyl Aldehyde
Modulates Enzymatic Methylation for Syringyl Monolignol Formation, a New View
of Monolignol Biosynthesis in Angiosperms. The Journal of biological chemistry
2000, 275, 6537–45.
(139) Osakabe, K.; Tsao, C. C.; Li, L.; Popko, J. L.; Umezawa, T.; Carraway, D. T.;
Smeltzer, R. H.; Joshi, C. P.; Chiang, V. L. Coniferyl aldehyde 5-hydroxylation and
methylation direct syringyl lignin biosynthesis in angiosperms. Proceedings of the
National Academy of Sciences of the United States of America 1999, 96, 8955–60.
(140) Higuchi, T. Look back over the studies of lignin biochemistry. Journal of Wood
Science 2006, 52, 2–8.
(141) Schoch, G.; Geopfert, S.; Morant, M.; Hehn, A.; Meyer Dulmann, P.; Werck-Reichart,
D. CYP98A3 from Arabidopsis thaliana is a 3′-hydroxylase of phenolic esters. A
missing link in the phenylpropanoid pathway. Journal of Biological Chemistry 2001,
276, 36566–36574.
(142) Ulbrich, B.; Zenk, M. H. Partial purification and properties of parahydroxycinnamoyl-CoA-shikimate-para-hydroxycinnamoyl transferase from higher
plants. Phytochemistry 1980, 19, 1625–1629.
(143) Hoffman, L.; Maury, S.; Martz, F.; Geoffrey, P.; Legrand, M. Purification, cloning and
properties of an acyltransferase controlling shikimate and quinate ester intermediates
in phenylpropanoid metabolism. Journal of Biological Chemistry 2002, 278, 95–103.
(144) Hu, W. J.; Kawaoka, a; Tsai, C. J.; Lung, J.; Osakabe, K.; Ebinuma, H.; Chiang, V. L.
Compartmentalized expression of two structurally and functionally distinct 4coumarate:CoA ligase genes in aspen (Populus tremuloides). Proceedings of the
National Academy of Sciences of the United States of America 1998, 95, 5407–12.
(145) Lauvergeat, V.; Lacomme, C.; Lacombe, E.; Lasserre, E.; Roby, D.; Grima-Pettenati,
J. Two cinnamoyl-CoA reductase (CCR) genes from Arabidopsis thaliana are
differentially expressed during development and in response to infection with
pathogenic bacteria. Phytochemistry 2001, 57, 1187–95.
133
(146) Steeves, V.; Förster, H.; Pommer, U.; Savidge, R. Coniferyl alcohol metabolism in
conifers -- I. Glucosidic turnover of cinnamyl aldehydes by UDPG: coniferyl alcohol
glucosyltransferase from pine cambium. Phytochemistry 2001, 57, 1085–93.
(147) Fagerstedt, K. V; Kukkola, E. M.; Koistinen, V. V. T.; Takahashi, J.; Marjamaa, K.
Cell wall lignin is polymerised by class III secretable plant peroxidases in Norway
spruce. Journal of integrative plant biology 2010, 52, 186–94.
(148) Freudenberg, K. Beirtrage zur Erforschung des Lignins. Angewandte Chemie 1956,
68, 508.
(149) Freudenberg, K. Biosynthesis and constitution of lignin. Nature 1959, 183, 1152–
1155.
(150) Tanahashi, M., Higuchi, T. Effect of the hydrophobic regions of hemicelluloses on
dehydrogenative polymerization of sinapyl alcohol. Mokuzai Gakkaishi 1990, 36,
424–428.
(151) Terashima, N, Atalla, R.H., Ralph, S.A., Landucci, L.L., Lapierre, C., Monties, B.
New preparation of lignin polymer models under conditions that approximate cell wall
lignification. Part 1. Holzforschung 1995, 49, 521–527.
(152) Terashima, N., Seguchi, Y. Heterogeneity in formation of lignin. IX. Factors
influencing the formation of condensed structures in lignins. Cellulose Chemistry and
Technology 1988, 22, 147–154.
(153) Guan, S.; Mlynar, J.; Sarkanen, S. S. Dehydrogenative polymerization of coniferyl
alcohol on macromolecular lignin templates. Phytochemistry 1997, 45.
(154) Donaldson, L. a Lignification and lignin topochemistry - an ultrastructural view.
Phytochemistry 2001, 57, 859–73.
(155) Davin, L. B.; Wang, H. B.; Crowell, a L.; Bedgar, D. L.; Martin, D. M.; Sarkanen, S.;
Lewis, N. G. Stereoselective bimolecular phenoxy radical coupling by an auxiliary
(dirigent) protein without an active center. Science 1997, 275, 362–6.
(156) Davin, L. B.; Lewis, N. G. Dirigent proteins and dirigent sites explain the mystery of
specificity of radical precursor coupling in lignan and lignin biosynthesis. Plant
physiology 2000, 123, 453–62.
(157) Davin, L. B.; Lewis, N. G. Lignin primary structures and dirigent sites. Current
opinion in biotechnology 2005, 16, 407–15.
(158) Chen, Y.-R.; Sarkanen, S. Macromolecular replication during lignin biosynthesis.
Phytochemistry 2010, 71, 453–62.
134
(159) Chen, Y.; Sarkanen, S. Macromolecular lignin replication : A mechanistic working
hypothesis. Phytochemistry Reviews 2003, 2, 235–255.
(160) Keller, B.; Templeton, M. D.; Lamb, C. J. Specific localization of a plant cell wall
glycine-rich protein in protoxylem cells of the vascular system. Proceedings of the
National Academy of Sciences of the United States of America 1989, 86, 1529–33.
(161) Ralph, J.; Grabber, J. H.; Hatfield, R. D. Lignin-ferulate cross-links in grasses: active
incorporation of ferulate polysaccharide esters into ryegrass lignins. Carbohydrate
Research 1995, 275, 167–178.
(162) Donaldson, L. A. Mechanical constraints on lignin deposition during lignification.
Wood Science and Technology 1994, 28, 111–118.
(163) Terashima, N.; Kitano, K.; Kojima, M.; Yoshida, M.; Yamamoto, H.; Westermark, U.
Nanostructural assembly of cellulose, hemicellulose, and lignin in the middle layer of
secondary wall of ginkgo tracheid. Journal of Wood Science 2009, 55, 409–416.
(164) Atalla, R. H.; Agarwal, U. P. Raman microprobe evidence for lignin orientation in the
cell walls of native woody tissue. Science (New York, N.Y.) 1985, 227, 636–8.
(165) Olsson, A.-M.; Bjurhager, I.; Gerber, L.; Sundberg, B.; Salmén, L. Ultra-structural
organisation of cell wall polymers in normal and tension wood of aspen revealed by
polarisation FTIR microspectroscopy. Planta 2011, 233, 1277–86.
(166) Åkerholm, M.; Salmén, L. The Oriented Structure of Lignin and its Viscoelastic
Properties Studied by Static and Dynamic FT-IR Spectroscopy. Holzforschung 2003,
57, 459–465.
(167) Lawoko, M.; Henriksson, G.; Gellerstedt, G. Structural differences between the lignincarbohydrate complexes present in wood and in chemical pulps. Biomacromolecules
2005, 6, 3467–73.
(168) Li, Jiebing; Martin-sampedro, Raquel, Pedrazzi, C.; Gellerstedt, G. Fractionation and
characterization of lignin-carbohydrate complexes (LCCs) from eucalyptus fibers.
Holzforschung 2011, 65, 43–50.
(169) Gratzl, J.S., Chen, C.-L. Chemistry of pulping: lignin reactions. In Lignin: Historical,
Biological, and Materials Perspectives.; Glasser, W.G., Northey, R.A., Schultz, T. P.,
Ed.; ACS Symposium Series 742; American Chemical Society: Washington, DC,
2000; pp. 392–421.
(170) Lai, Y.Z., Sarkanen, K. V. Isolation and structural studies. In. In Lignins: Occurrence,
formation, structure, and reactions; Sarkanen, K.V., Ludwig, C. ., Ed.; Wiley
Interscience: New York, 1971; pp. 165–240.
135
(171) Wallis, A. F. . Solvolysis by acids and bases. In Lignins: Occurrence, formation,
structure, and reactions; Sarkanen, K.V., Ludwig, C. H., Ed.; Wiley Interscience:
New York, 1971; pp. 345–372.
(172) Glennie, D. W. Reactions in sulfite pulping. In Lignins: Occurrence, formation,
structure, and reactions; Sarkanen, K.V., Ludwig, C. H., Ed.; Wiley Interscience:
New York, 1971; pp. 597–631.
(173) Marton, J. Reactions in alkaline pulping. In Lignins: Occurrence, formation, structure,
and reactions; Sarkanen, K.V., Ludwig, C. H., Ed.; Wiley Interscience: New York,
1971; pp. 639–694.
(174) Sarkanen, K. V. Chemistry of solvent pulping. Tappi Journal 1990, 73, 215–219.
(175) McDonough, T. J. The chemistry of organosolv delignification. Tappi Journal 1993,
76, 186–193.
(176) Lora, J.H., Wu, C.F., Pye, K., Balatinecz, J. J. Characteristics and potential
applications of lignin produced by an organosolv process. In Lignin: Properties and
Materials; Glasser, W.G., Sarkanen, S., Ed.; ACS Symposium Series 397, American
Chemical Society, 1989; pp. 312–323.
(177) Scholze, B.; Meier, D. Characterization of the water-insoluble fraction from pyrolysis
oil (pyrolytic lignin). Part I. PY–GC/MS, FTIR, and functional groups. Journal of
Analytical and Applied Pyrolysis 2001, 60, 41–54.
(178) Bayerbach, R.; Nguyen, V. D.; Schurr, U.; Meier, D. Characterization of the waterinsoluble fraction from fast pyrolysis liquids (pyrolytic lignin). Journal of Analytical
and Applied Pyrolysis 2006, 77, 95–101.
(179) Gierer, J. Chemistry of delignification. Wood Science and Technology 1985, 19, 289–
312.
(180) Fredheim, G. E.; Braaten, S. M.; Christensen, B. E. Molecular weight determination of
lignosulfonates by size-exclusion chromatography and multi-angle laser light
scattering. Journal of chromatography. A 2002, 942, 191–9.
(181) Braaten, S. M.; Christensen, B. E.; Fredheim, G. E. Comparison of Molecular Weight
and Molecular Weight Distributions of Softwood and Hardwood Lignosulfonates.
Journal of Wood Chemistry and Technology 2003, 23, 197–215.
(182) Ekeberg, D.; Gretland, K. S.; Gustafsson, J.; Bråten, S. M.; Fredheim, G. E.
Characterisation of lignosulphonates and kraft lignin by hydrophobic interaction
chromatography. Analytica Chimica Acta 2006, 565, 121–128.
136
(183) Lora, J. H.; Glasser, W. G. Recent Industrial Applications of Lignin : A Sustainable
Alternative to Nonrenewable Materials. Journal of Polymers and the Environment
2002, 10, 39–48.
(184) Young, R. a. Comparison of the properties of chemical cellulose pulps. Cellulose
1994, 1, 107–130.
(185) Mabee, W.E., Gregg, D.J., Arato, C., Berlin, A., Bura, R. Gilkes, N., Mirochnik, O.,
Pan, X., Pye, E.K., Saddler, J. N. Updates on Softwood-to-Ethanol Process. Applied
Biochemistry and Biotechnology 2006, 129-132, 55–70.
(186) Pan, X.; Arato, C.; Gilkes, N.; Gregg, D.; Mabee, W.; Pye, K.; Xiao, Z.; Zhang, X.;
Saddler, J. Biorefining of softwoods using ethanol organosolv pulping: preliminary
evaluation of process streams for manufacture of fuel-grade ethanol and co-products.
Biotechnology and bioengineering 2005, 90, 473–81.
(187) Pan, X.; Xie, D.; Yu, R. W.; Lam, D.; Saddler, J. N. Pretreatment of Lodgepole Pine
Killed by Mountain Pine Beetle Using the Ethanol Organosolv Process: Fractionation
and Process Optimization. Industrial & Engineering Chemistry Research 2007, 46,
2609–2617.
(188) Pan, X.; Xie, D.; Yu, R. W.; Saddler, J. N. The bioconversion of mountain pine beetlekilled lodgepole pine to fuel ethanol using the organosolv process. Biotechnology and
bioengineering 2008, 101, 39–48.
(189) Schuerch, C. J. The Solvent Properties of Liquids and Their Relation to the Solubility ,
Swelling , Isolation and Fractionation of Lignin. Journal of the American Chemical
Society 1952, 74, 5061–5067.
(190) Paszner, L., Behera, N. C. Topochemistry of softwood delignification by alkali earth
metal salt catalyzed organosolv pulping. Holzforschung 1989, 43, 159–168.
(191) Norgren, M.; Lindström, B. Physico-Chemical Characterization of a Fractionated
Kraft Lignin. Holzforschung 2000, 54, 528–534.
(192) Kadla, J. F.; Kubo, S. Miscibility and Hydrogen Bonding in Blends of Poly(ethylene
oxide) and Kraft Lignin. Macromolecules 2003, 36, 7803–7811.
(193) Guigo, N.; Mija, A.; Vincent, L.; Sbirrazzuoli, N. Molecular mobility and relaxation
process of isolated lignin studied by multifrequency calorimetric experiments.
Physical chemistry chemical physics : PCCP 2009, 11, 1227–36.
(194) Toledano, a.; García, a.; Mondragon, I.; Labidi, J. Lignin separation and fractionation
by ultrafiltration. Separation and Purification Technology 2010, 71, 38–43.
137
(195) Wallberg, O.; Jönsson, A.-S.; Wimmerstedt, R. Fractionation and concentration of
kraft black liquor lignin with ultrafiltration. Desalination 2003, 154, 187–199.
(196) Toledano, a.; Serrano, L.; Garcia, a.; Mondragon, I.; Labidi, J. Comparative study of
lignin fractionation by ultrafiltration and selective precipitation. Chemical Engineering
Journal 2010, 157, 93–99.
(197) Morck, R., Yoshida, H., Kringstad, K.P., Hatakeyama, H. Fractionation of Kraft lignin
by successive extraction with organic solvents I. Functional groups, 13C-NMR spectra
and molecular weight distributions. Holzforschung 1986, 40 (Suppl., 51–60.
(198) Yoshida, H., Morck, R., Kringstad, K.P., Hatakeyama, H. Fractionation of Kraft
Lignin by Successive Extraction with Organic Solvents. II. Thermal Properties of
Kraft Lignin Fractions. Holzforschung 1987, 41, 171–176.
(199) Vanderlaan, M.N.; Thring, R. W. Polyurethanes from Alcell lignin fractions obtained
by sequential solvent extraction. Biomass and Bioenergy 1998, 14, 525–531.
(200) Sun, R.; Tomkinson, J.; Griffiths, S. Fractional and Physico- Chemical Analysis of
Soda-AQ Lignin by Successive Extraction with Organic Solvents from Oil Palm EFB
Fiber. International Journal of Polymer Analysis and Characterization 2000, 5, 531–
547.
(201) Wang, K.; Xu, F.; Sun, R. Molecular Characteristics of Kraft-AQ Pulping Lignin
Fractionated by Sequential Organic Solvent Extraction. International journal of
molecular sciences 2010, 11, 2988–3001.
(202) Gosselink, R.J.A., van Dam, J.E.G., de Jong, E., Scott, E.L., Sanders, J.P.M., Li, J.,
Gellerstedt, G. Fractionation , analysis , and PCA modeling of properties of four
technical lignins for prediction of their application potential in binders. Holzforschung
2010, 64, 193–200.
(203) Ropponen, J.; Ra, L.; Rovio, S.; Mikkonen, H.; van de Pas, D.; Tamminen, T. Solvent
extraction as a means of preparing homogeneous lignin fractions. Holzforschung 2011,
65, 543–549.
(204) Formhals, A. Process and apparatus for preparing artificial threads. U.S. Patent No.
1,975,504 1934, 1934.
(205) Taylor, G. Disintegration of Water Drops in an Electric Field. Proceedings of the
Royal Society A: Mathematical, Physical and Engineering Sciences 1964, 280, 383–
397.
(206) Taylor, G. I.; McEwan, A. D. The stability of a horizontal fluid interface in a vertical
electric field. Journal of Fluid Mechanics 1965, 22, 1–15.
138
(207) Taylor, G. Electrically driven jets. Proc. Roy. Soc. Lond. A 1969, 313, 453–475.
(208) Baumgarten, P. K. Electrostatic spinning of acrylic microfibers. Journal of Colloid
and Interface Science 1971, 36, 71–79.
(209) Doshi, J.; Reneker, D. H. Electrospinning process and applications of electrospun
fibers. Journal of Electrostatics 1995, 35, 151–160.
(210) Srinivasad, G., Reneker, D. H. Structure and morphology of small diameter
electrospun aramid fibers. Polymer International 1995, 36, 195–201.
(211) Thavasi, V.; Singh, G.; Ramakrishna, S. Electrospun nanofibers in energy and
environmental applications. Energy & Environmental Science 2008, 1, 205.
(212) Luo, C. J.; Stoyanov, S. D.; Stride, E.; Pelan, E.; Edirisinghe, M. Electrospinning
versus fibre production methods: from specifics to technological convergence.
Chemical Society reviews 2012, 41, 4708–35.
(213) Reneker, D. H.; Yarin, A. L.; Fong, H.; Koombhongse, S. Bending instability of
electrically charged liquid jets of polymer solutions in electrospinning. Journal of
Applied Physics 2000, 87, 4531.
(214) Yarin, a. L.; Koombhongse, S.; Reneker, D. H. Bending instability in electrospinning
of nanofibers. Journal of Applied Physics 2001, 89, 3018.
(215) Yarin, a. L.; Koombhongse, S.; Reneker, D. H. Taylor cone and jetting from liquid
droplets in electrospinning of nanofibers. Journal of Applied Physics 2001, 90, 4836.
(216) Hohman, M. M.; Shin, M.; Rutledge, G.; Brenner, M. P. Electrospinning and
electrically forced jets. I. Stability theory. Physics of Fluids 2001, 13, 2201.
(217) Hohman, M. M.; Shin, M.; Rutledge, G.; Brenner, M. P. Electrospinning and
electrically forced jets. II. Applications. Physics of Fluids 2001, 13, 2221.
(218) Feng, J. J. The stretching of an electrified non-Newtonian jet: A model for
electrospinning. Physics of Fluids 2002, 14, 3912.
(219) Feng, J. J. Stretching of a straight electrically charged viscoelastic jet. Journal of NonNewtonian Fluid Mechanics 2003, 116, 55–70.
(220) Spivak, a. F.; Dzenis, Y. a. Asymptotic decay of radius of a weakly conductive viscous
jet in an external electric field. Applied Physics Letters 1998, 73, 3067.
(221) Spivak, A. E.; Dzenis, Y. A. I.; Reneker, D. H. A model of steady state jet in the
electrospinning process. Mechanics Research Communications 2000, 27, 37–42.
139
(222) Carroll, C. P.; Joo, Y. L. Electrospinning of viscoelastic Boger fluids: Modeling and
experiments. Physics of Fluids 2006, 18, 053102.
(223) Carroll, C. P.; Joo, Y. L. Axisymmetric instabilities of electrically driven viscoelastic
jets. Journal of Non-Newtonian Fluid Mechanics 2008, 153, 130–148.
(224) Carroll, C. P.; Joo, Y. L. Axisymmetric instabilities in electrospinning of highly
conducting, viscoelastic polymer solutions. Physics of Fluids 2009, 21, 103101.
(225) Helgeson, M. E.; Grammatikos, K. N.; Deitzel, J. M.; Wagner, N. J. Theory and
kinematic measurements of the mechanics of stable electrospun polymer jets. Polymer
2008, 49, 2924–2936.
(226) Han, T.; Yarin, A. L.; Reneker, D. H. Viscoelastic electrospun jets: Initial stresses and
elongational rheometry. Polymer 2008, 49, 1651–1658.
(227) Kakade, M. V; Givens, S.; Gardner, K.; Lee, K. H.; Chase, D. B.; Rabolt, J. F. Electric
field induced orientation of polymer chains in macroscopically aligned electrospun
polymer nanofibers. Journal of the American Chemical Society 2007, 129, 2777–82.
(228) Fridrikh, S.; Yu, J.; Brenner, M.; Rutledge, G. Controlling the Fiber Diameter during
Electrospinning. Physical Review Letters 2003, 90, 1–4.
(229) Yu, J.; Fridrikh, S.; Rutledge, G. The role of elasticity in the formation of electrospun
fibers. Polymer 2006, 47, 4789–4797.
(230) McKee, M. G.; Wilkes, G. L.; Colby, R. H.; Long, T. E. Correlations of Solution
Rheology with Electrospun Fiber Formation of Linear and Branched Polyesters.
Macromolecules 2004, 37, 1760–1767.
(231) Shenoy, S.; Bates, W.; Frisch, H.; Wnek, G. Role of chain entanglements on fiber
formation during electrospinning of polymer solutions: good solvent, non-specific
polymer?polymer interaction limit. Polymer 2005, 46, 3372–3384.
(232) Thompson, C. J.; Chase, G. G.; Yarin, a. L.; Reneker, D. H. Effects of parameters on
nanofiber diameter determined from electrospinning model. Polymer 2007, 48, 6913–
6922.
(233) Theron, S. a.; Zussman, E.; Yarin, a. L. Experimental investigation of the governing
parameters in the electrospinning of polymer solutions. Polymer 2004, 45, 2017–2030.
(234) Son, W. K.; Youk, J. H.; Lee, T. S.; Park, W. H. The effects of solution properties and
polyelectrolyte on electrospinning of ultrafine poly(ethylene oxide) fibers. Polymer
2004, 45, 2959–2966.
140
(235) Casper, C. L.; Stephens, J. S.; Tassi, N. G.; Chase, D. B.; Rabolt, J. F. Controlling
Surface Morphology of Electrospun Polystyrene Fibers: Effect of Humidity and
Molecular Weight in the Electrospinning Process. Macromolecules 2004, 37, 573–
578.
(236) Tan, S.-H.; Inai, R.; Kotaki, M.; Ramakrishna, S. Systematic parameter study for ultrafine fiber fabrication via electrospinning process. Polymer 2005, 46, 6128–6134.
(237) De Gennes, P.-G. Scaling Concepts in Polymer Physics; Cornell University Press:
Ithaca and London, 1979.
(238) Colby, R. H.; Fetters, L. J.; Funk, W. G.; Graessley, W. W. Effects of Concentration
and Thermodynamic Interaction on the viscoelastic properties of polymer solutions.
Macromolecules 1991, 24, 3873–3882.
(239) Macosko, C. W. Rheology principles, measurements, and applications; Wiley-VCH:
New York, 1994.
(240) McKee, M. G.; Elkins, C. L.; Long, T. E. Influence of self-complementary hydrogen
bonding on solution rheology/electrospinning relationships. Polymer 2004, 45, 8705–
8715.
(241) McKee, M. G.; Layman, J. M.; Cashion, M. P.; Long, T. E. Phospholipid nonwoven
electrospun membranes. Science 2006, 311, 353–5.
(242) Lallave, M.; Bedia, J.; Ruiz-Rosas, R.; Rodríguez-Mirasol, J.; Cordero, T.; Otero, J.
C.; Marquez, M.; Barrero, a.; Loscertales, I. G. Filled and Hollow Carbon Nanofibers
by Coaxial Electrospinning of Alcell Lignin without Binder Polymers. Advanced
Materials 2007, 19, 4292–4296.
(243) Ruiz-Rosas, R.; Bedia, J.; Lallave, M.; Loscertales, I. G.; Barrero, a.; RodríguezMirasol, J.; Cordero, T. The production of submicron diameter carbon fibers by the
electrospinning of lignin. Carbon 2010, 48, 696–705.
(244) Ago, M.; Okajima, K.; Jakes, J. E.; Park, S.; Rojas, O. J. Lignin-based electrospun
nanofibers reinforced with cellulose nanocrystals. Biomacromolecules 2012, 13, 918–
26.
(245) Seo, D. K.; Jeun, J. P.; Kim, H. Bin; Kang, P. H. Preparation and characterization of
the carbon nanofiber mat produced from electrospun PAN/lignin precursors by
electron beam irradiation. Rev. Adv. Mater. Sci. 2011, 28, 31–34.
(246) Chang, L.C.; Read, T. A. Plastic deformation and diffusionless phase changes in
metals: the gold-cadmium beta phase. Trans AIME 1951, 189, 47.
141
(247) Buehler, W. J.; Gilfrich, J. V.; Wiley, R. C. Effect of Low-Temperature Phase
Changes on the Mechanical Properties of Alloys near Composition TiNi. Journal of
Applied Physics 1963, 34, 1475.
(248) Vernon, L.B.; Vernon, H. M. Producing Molded Articles such as Dentures from
Thermoplastic Synthetic Resins. US Patent 2234993 1941.
(249) Rainer, W.C.; Redding, E.M.; Hitov, J.J.; Sloan A.W.; Stewart, W. D. Heat-shrinkable
Polyethylene. US Patent 3144398 1964.
(250) Hayashi, S. Technical report on preliminary investigation of shape memory polymers,
Nagoya research and development center, Mitsubishi Heavy Industries Inc, 1990.;
1990.
(251) Zugenmaier, P. Contribution to the historical development of macromolecular
chemistry - exemplified on cellulose. Cellulose Chemistry and Technology 2009, 43,
351–378.
(252) Salmén, L.; Burgert, I. Cell wall features with regard to mechanical performance. A
review. Holzforschung 2009, 63, 121–129.
(253) Chowdhury, S.; Madsen, L. a; Frazier, C. E. Probing alignment and phase behavior in
intact wood cell walls using 2H NMR spectroscopy. Biomacromolecules 2012, 13,
1043–50.
(254) Keckes, J.; Burgert, I.; Frühmann, K.; Müller, M.; Kölln, K.; Hamilton, M.;
Burghammer, M.; Roth, S. V; Stanzl-Tschegg, S.; Fratzl, P. Cell-wall recovery after
irreversible deformation of wood. Nature materials 2003, 2, 810–4.
(255) Hill, S. J.; Franich, R. A.; Callaghan, P. T.; Newman, R. H. Nature’s Nanocomposites:
A New Look at Molecular Architecture in Wood Cell Walls. New Zealand Journal of
Forestry Science 2009, 39, 251–257.
(256) Altaner, C. M.; Jarvis, M. C. Modelling polymer interactions of the “molecular
Velcro” type in wood under mechanical stress. Journal of theoretical biology 2008,
253, 434–45.
(257) Kubo, S.; Kadla, J. F. Hydrogen bonding in lignin: a Fourier transform infrared model
compound study. Biomacromolecules 2005, 6, 2815–21.
(258) Deng, Y.; Feng, X.; Zhou, M.; Qian, Y.; Yu, H.; Qiu, X. Investigation of aggregation
and assembly of alkali lignin using iodine as a probe. Biomacromolecules 2011, 12,
1116–25.
(259) Guerra, A.; Gaspar, A. R.; Contreras, S.; Lucia, L. a; Crestini, C.; Argyropoulos, D. S.
On the propensity of lignin to associate: a size exclusion chromatography study with
142
lignin derivatives isolated from different plant species. Phytochemistry 2007, 68,
2570–83.
(260) Sarkanen, S.; Teller, D. C.; Stevens, C. R.; McCarthy, J. L. Lignin. 20. Associative
interactions between kraft lignin components. Macromolecules 1984, 17, 2588–2597.
(261) Page, D.H., Seth, R.S. Jordan, B.D., Barbe, M. C. Curls, crimps, kinks, and
microcompressions in pulp fibres- their origin, measurement, and significance.
Proceedings 8th Fundamental Research Symposium 1985.
(262) Perepelkin, K. E. Carbon fibres with specific physical and physicochemical properties
based on hydrated cellulose and polyacrylonitrile precursors. A review. Fibre
Chemistry 2002, 34, 271–280.
(263) Edie, D. D. The effect of processing on the structure and properties of carbon fibers.
Carbon 1998, 36, 345–362.
(264) Sheehan, J. E.; Buesking, K. W.; Sullivan, B. J. Carbon-Carbon Composites. Annual
Review of Materials Science 1994, 24, 19–44.
(265) Aber, S.; Khataee, A.; Sheydaei, M. Optimization of activated carbon fiber preparation
from Kenaf using K2HPO4 as chemical activator for adsorption of phenolic
compounds. Bioresource technology 2009, 100, 6586–91.
(266) Avraham, E.; Yaniv, B.; Soffer, A.; Aurbach, D. Developing Ion Electroadsorption
Stereoselectivity , by Pore Size Adjustment with Chemical Vapor Deposition onto
Active Carbon Fiber Electrodes . Case of Ca 2 + / Na + Separation in Water
Capacitive Desalination. Journal of Physical Chemistry C 2008, 112, 7385–7389.
(267) Mochida, I.; Shirahama, N.; Kawano, S.; Korai, Y.; Yasutake, A.; Tanoura, M.; Fujii,
S.; Yoshikawa, M. NO oxidation over activated carbon fiber ( ACF ). Part 1 .
Extended kinetics over a pitch based ACF of very large surface area. Fuel 2000, 79,
1713–1723.
(268) Phan, N. H.; Rio, S.; Faur, C.; Le Coq, L.; Le Cloirec, P.; Nguyen, T. H. Production of
fibrous activated carbons from natural cellulose (jute, coconut) fibers for water
treatment applications. Carbon 2006, 44, 2569–2577.
(269) Sanchez-Montero, M.J.; Salvador, F.; Izquierdo, C. Reactivity and Porosity of a
Carbon Fiber Activated with Supercritical CO 2. Journal of Physical Chemistry C
2008, 112, 4991–4999.
(270) Lin, J.-H.; Ko, T.-H.; Lin, Y.-H.; Pan, C.-K. Various Treated Conditions to Prepare
Porous Activated Carbon Fiber for Application in Supercapacitor Electrodes. Energy
& Fuels 2009, 23, 4668–4677.
143
(271) Koga, H.; Umemura, Y.; Ishihara, H.; Kitaoka, T.; Tomoda, A.; Suzuki, R.; Wariishi,
H. Paper-structured fiber composites impregnated with platinum nanoparticles
synthesized on a carbon fiber matrix for catalytic reduction of nitrogen oxides. Applied
Catalysis B: Environmental 2009, 90, 699–704.
(272) Hung, C.-M. Activity of Cu-activated carbon fiber catalyst in wet oxidation of
ammonia solution. Journal of hazardous materials 2009, 166, 1314–20.
(273) Chiang, Y.-C.; Lee, C.-C.; Lee, H.-C. Characterization of microstructure and surface
properties of heat-treated PAN-and rayon-based activated carbon fibers. Journal of
Porous Materials 2006, 14, 227–237.
(274) Mochida, I.; Yoon, S.-H.; Korai, Y. Mesoscopic structure and properties of liquid
crystalline mesophase pitch and its transformation into carbon fiber. Chemical record
2002, 2, 81–101.
(275) Liu, C.-L.; Dong, W.-S.; Song, J.-R.; Liu, L. Evolution of microstructure and
properties of phenolic fibers during carbonization. Materials Science and Engineering:
A 2007, 459, 347–354.
(276) Qiao, W. M.; Huda, M.; Song, Y.; Yoon, S.; Korai, Y.; Mochida, I. Carbon Fibers and
Films Based on Biomass Resins. Energy and Fuels 2005, 19, 2576–2582.
(277) Rahaman, M. S. a.; Ismail, a. F.; Mustafa, a. A review of heat treatment on
polyacrylonitrile fiber. Polymer Degradation and Stability 2007, 92, 1421–1432.
(278) Liu, F.; Wang, H.; Xue, L.; Fan, L.; Zhu, Z. Effect of microstructure on the
mechanical properties of PAN-based carbon fibers during high-temperature
graphitization. Journal of Materials Science 2008, 43, 4316–4322.
(279) Radishevskii, M. B.; Serkov, A. T.; Budnitskii, G. A.; Medvedev, V. A.; Zlatoustova,
L. A. Improvement of manufacturing technology for ultrastrong and high-modulus
carbon fibres. Fibre Chemistry 2005, 37, 327–331.
(280) M60J Data sheet, Toray industries http://www.toraycfa.com/.
(281) Gribanov, A. V; Sazanov, Y. N. Polyacrylonitrile : Carbonization Problems. Russian
Journal of Applied Chemistry 2008, 81, 919–932.
(282) Jain, M.; Abhiraman, A. S. Conversion of acrylonitrile-based precursor fibres to
carbon fibres Part 1 A review of the physical and morphological aspects. Journal of
Materials Science 1987, 22, 278–300.
(283) Ji, M.; Wang, C.; Bai, Y.; Yu, M.; Wang, Y. Structural Evolution of Polyacrylonitrile
Precursor Fibers during Preoxidation and Carbonization. Polymer Bulletin 2007, 59,
527–536.
144
(284) Liu, J.; Zhou, P.; Zhang, L.; Ma, Z.; Liang, J.; Fong, H. Thermo-chemical reactions
occurring during the oxidative stabilization of electrospun polyacrylonitrile precursor
nanofibers and the resulting structural conversions. Carbon 2009, 47, 1087–1095.
(285) Zhang, W.; Liu, J.; Wu, G. Evolution of structure and properties of PAN precursors
during their conversion to carbon fibers. Carbon 2003, 41, 2805–2812.
(286) T1000G Data sheet, Toray Industries.
(287) Otani, S.; Fukuoka, Y.; Igarashi, B.; Sasaki, K. Method for Producing Carbonized
Lignin Fiber. US Patent 3,461,082. 1969.
(288) Sudo, K.; Shimizu, K.; Nakashima, N.; Yokoyama, A. A new modification method of
exploded lignin for the preparation of a carbon-fiber precursor. Journal of Applied
Polymer Science 1993, 48, 1485–91.
(289) Uraki, Y., Kubo, S., Nigo, N., Sano, Y., Sasaya, T. Preparation of carbon-fibers from
organosolv lignin obtained by aqueous acetic-acid pulping. Holzforschung 1995, 49,
343–350.
(290) Kubo S, Ishikawa N, Uraki Y, S. Y. Preparation of lignin fibers from softwood acetic
acid lignin—Relationship between fusibility and the chemical structure of lignin.
Mokuzai Gakkaishi 1997, 43, 655–62.
(291) Kubo, S.; Uraki, Y.; Sano, Y. Preparation of carbon fibers from softwood lignin by
atmospheric acetic acid pulping. Carbon 1998, 36, 1119–1124.
(292) Dave, V.; Prasad, A.; Marand, H.; Glasser, W. G. Molecular organization of lignin
during carbonization. Polymer 1993, 34, 3144–3154.
(293) Johnson, D. J.; Tomizuka, I.; Watanabe, O. The fine structure of lignin-based carbon
fibres. Carbon 1975, 13, 321–325.
(294) Kubo, S.; Kadla, J. F. Kraft lignin/poly(ethylene oxide) blends: Effect of lignin
structure on miscibility and hydrogen bonding. Journal of Applied Polymer Science
2005, 98, 1437–1444.
(295) Kubo, S.; Kadla, J. F. Poly(Ethylene Oxide)/Organosolv Lignin Blends: Relationship
between Thermal Properties, Chemical Structure, and Blend Behavior.
Macromolecules 2004, 37, 6904–6911.
(296) Kubo, S.; Kadla, J. F. Effect of poly(ethylene oxide) molecular mass on miscibility
and hydrogen bonding with lignin. Holzforschung 2006, 60, 245–252.
145
(297) Capanema, E. a; Balakshin, M. Y.; Kadla, J. F. A comprehensive approach for
quantitative lignin characterization by NMR spectroscopy. Journal of agricultural and
food chemistry 2004, 52, 1850–60.
(298) Bayat, M.; Yang, H.; Ko, F. Electromagnetic properties of electrospun Fe3O4/carbon
composite nanofibers. Polymer 2011, 52, 1645–1653.
(299) Stelter, M.; Brenn, G.; Yarin, A.L.; Durst, F. Validation and application of a novel
elongational device. Journal of Rheology 2000, 44, 595–616.
(300) Stelter, M.; Brenn, G.; Yarin, a. L.; Singh, R. P.; Durst, F. Investigation of the
elongational behavior of polymer solutions by means of an elongational rheometer.
Journal of Rheology 2002, 46, 507.
(301) Anna, S. L.; McKinley, G. H. Elasto-capillary thinning and breakup of model elastic
liquids. Journal of Rheology 2001, 45, 115.
(302) Rodd, L. E.; Scott, T. P.; Cooper-white, J. J.; McKinley, G. H. Capillary Break-up
Rheometry of Low-Viscosity Elastic Fluids. Applied Rheology 2005, 15, 12–27.
(303) Entov, V. M.; Hinch, E. J. Effect of a spectrum of relaxation times on the capillary
thinning of a filament of elastic liquid. Journal of Non-Newtonian Fluid Mechanics
1997, 72, 31–53.
(304) Kolte, M. I.; Szabo, P. Capillary thinning of polymeric filaments. Journal of Rheology
1999, 43, 609–625.
(305) Dong, D.; Fricke, a Intrinsic viscosity and the molecular weight of kraft lignin.
Polymer 1995, 36, 2075–2078.
(306) Ferrari, a.; Robertson, J. Interpretation of Raman spectra of disordered and amorphous
carbon. Physical Review B 2000, 61, 14095–14107.
(307) Ferrari, a.; Robertson, J. Resonant Raman spectroscopy of disordered, amorphous, and
diamondlike carbon. Physical Review B 2001, 64, 1–13.
(308) Colby, R. H.; Rubinstein, M. Two-parameter scaling for polymers in Θ solvents.
Macromolecules 1990, 23, 2753–2757.
(309) Garver, T.M.; Callaghan, P. T. Hydrodynamics of Kraft Lignins. Macromolecules
1991, 24, 420–30.
(310) Sarkanen, S.; Teller, D. C.; Abramowski, E.; McCarthy, J. L. Lignin. 19. Kraft lignin
component conformation and associated complex configuration in aqueous alkaline
solution. Macromolecules 1982, 15, 1098–1104.
146
(311) Anna, S. L.; McKinley, G. H.; Nguyen, D. a.; Sridhar, T.; Muller, S. J.; Huang, J.;
James, D. F. An interlaboratory comparison of measurements from filament-stretching
rheometers using common test fluids. Journal of Rheology 2001, 45, 83.
(312) Connors, W.J.; Sarkanen, S.; McCarthy, J. L. Gel chromatography and association
complexes of lignin. Holzforschung 1980, 34, 80–85.
(313) Lindstrom, T. The colloidal behaviour of kraft lignin Part I .: Association and gelation
of kraft lignin in aqueous solutions. Colloid and Polymer Science 1979, 257, 277–285.
(314) Clasen, C.; Plog, J. P.; Kulicke, W.-M.; Owens, M.; Macosko, C.; Scriven, L. E.;
Verani, M.; McKinley, G. H. How dilute are dilute solutions in extensional flows?
Journal of Rheology 2006, 50, 849.
(315) Tirtaatmadja, V.; McKinley, G. H.; Cooper-White, J. J. Drop formation and breakup
of low viscosity elastic fluids: Effects of molecular weight and concentration. Physics
of Fluids 2006, 18, 043101.
(316) Takahashi, T.; Takimoto, J.; Koyama, K. Elongational Viscosity for Miscible and
Immiscible Polymer Blends. II. Blends with a Small Amount of UHMW Polymer.
Journal of Applied Polymer Science 1999, 72, 961–969.
(317) Starý, Z.; Pemsel, T.; Baldrian, J.; Münstedt, H. Influence of a compatibilizer on the
morphology development in polymer blends under elongation. Polymer 2012, 53,
1881–1889.
(318) Regev, O.; Vandebril, S.; Zussman, E.; Clasen, C. The role of interfacial
viscoelasticity in the stabilization of an electrospun jet. Polymer 2010, 51, 2611–2620.
(319) Braun, J. L.; Holtman, K. M.; Kadla, J. F. Lignin-based carbon fibers: Oxidative
thermostabilization of kraft lignin. Carbon 2005, 43, 385–394.
(320) Nordström, Y.; Norberg, I.; Sjöholm, E.; Drougge, R. A new softening agent for melt
spinning of softwood kraft lignin. Journal of Applied Polymer Science 2012, doi:
10.1002/app.38795
(321) Uraki, Y.; Sugiyama, Y.; Koda, K.; Kubo, S.; Kishimoto, T.; Kadla, J. F. Thermal
mobility of β-O-4-type artificial lignin. Biomacromolecules 2012, 13, 867–72.
(322) Kringstad, K.P.; Morck, R. 13C-NMR spectra of Kraft lignins. Holzforschung 1983,
37, 237–244.
(323) Liitiä, T. M.; Maunu, S. L.; Hortling, B.; Toikka, M.; Kilpeläinen, I. Analysis of
technical lignins by two- and three-dimensional NMR spectroscopy. Journal of
agricultural and food chemistry 2003, 51, 2136–43.
147
(324) Cathala, B.; Saake, B.; Faix, O.; Monties, B. Association behaviour of lignins and
lignin model compounds studied by multidetector size-exclusion chromatography.
Journal of Chromatography A 2003, 1020, 229–239.
(325) Li, Y.; Mlynar, J.; Sarkanen, S. The First 85 % Kraft Lignin-Based Thermoplastics.
Journal of Polymer Science Part B: Polymer Physics 1997, 35, 1899–1910.
(326) Li, Y.; Sarkanen, S. Alkylated Kraft Lignin-Based Thermoplastic Blends with
Aliphatic Polyesters. Macromolecules 2002, 35, 9707–9715.
(327) Kidoaki, S.; Kwon, I. K.; Matsuda, T. Structural features and mechanical properties of
in situ-bonded meshes of segmented polyurethane electrospun from mixed solvents.
Journal of biomedical materials research. Part B, Applied biomaterials 2006, 76,
219–29.
(328) Lee, K. H.; Kim, H. Y.; Ryu, Y. J.; Kim, K. W.; Choi, S. W. Mechanical behavior of
electrospun fiber mats of poly(vinyl chloride)/polyurethane polyblends. Journal of
Polymer Science Part B: Polymer Physics 2003, 41, 1256–1262.
(329) Dharmadhikary, R. K.; Davis, H.; Gilmore, T. F.; Batra, S. K. Influence of Fiber
Structure Properties of Thermally point bonded polypropylene nonwovens. Textile
Research Journal 1999, 69, 725–734.
(330) Michielsen, S.; Pourdeyhimi, B.; Desai, P. Review of thermally point-bonded
nonwovens: Materials, processes, and properties. Journal of Applied Polymer Science
2006, 99, 2489–2496.
(331) Rawal, A. A Modified Micromechanical Model for the Prediction of Tensile Behavior
of Nonwoven Structures. Journal of Industrial Textiles 2006, 36, 133–149.
(332) Silberstein, M. N.; Pai, C.; Rutledge, G. C.; Boyce, M. C. Journal of the Mechanics
and Physics of Solids Elastic – plastic behavior of non-woven fibrous mats. Journal of
the Mechanics and Physics of Solids 2012, 60, 295–318.
(333) Imaizumi, S.; Matsumoto, H.; Suzuki, K.; Minagawa, M.; Kimura, M.; Tanioka, A.
Phenolic Resin-Based Carbon Thin Fibers Prepared by Electrospinning: Additive
Effects of Poly(vinyl butyral) and Electrolytes. Polymer Journal 2009, 41, 1124–1128.
(334) Park, S. H.; Kim, C.; Yang, K. S. Preparation of carbonized fiber web from
electrospinning of isotropic pitch. Synthetic Metals 2004, 143, 175–179.
(335) Pimenta, M. a; Dresselhaus, G.; Dresselhaus, M. S.; Cançado, L. G.; Jorio, a; Saito, R.
Studying disorder in graphite-based systems by Raman spectroscopy. Physical
chemistry chemical physics : PCCP 2007, 9, 1276–91.
148
(336) Shimodaira, N.; Masui, a. Raman spectroscopic investigations of activated carbon
materials. Journal of Applied Physics 2002, 92, 902.
(337) Jawhari, T.; Roid, A.; Casado, J. Raman spectroscopic characterization of some
commercially available carbon black materials. Carbon 1995, 33, 1561–1565.
(338) Kim, C.; Park, S.-H.; Cho, J.-I.; Lee, D.-Y.; Park, T.-J.; Lee, W.-J.; Yang, K.-S.
Raman spectroscopic evaluation of polyacrylonitrile-based carbon nanofibers prepared
by electrospinning. Journal of Raman Spectroscopy 2004, 35, 928–933.
(339) Lee, S. Structural Evolution of Polyacrylonitrile Fibers in Stabilization and
Carbonization. Advances in Chemical Engineering and Science 2012, 02, 275–282.
(340) Ishimaru, K.; Hata, T.; Bronsveld, P.; Meier, D.; Imamura, Y. Spectroscopic analysis
of carbonization behavior of wood, cellulose and lignin. Journal of Materials Science
2006, 42, 122–129.
(341) Dallmeyer, I.; Ko, F.; Kadla, J. F. Electrospinning of Technical Lignins for the
Production of Fibrous Networks. Journal of Wood Chemistry and Technology 2010,
30, 315–329.
(342) Arinstein, A.; Burman, M.; Gendelman, O.; Zussman, E. Effect of supramolecular
structure on polymer nanofibre elasticity. Nature nanotechnology 2007, 2, 59–62.
(343) Naraghi, M.; Arshad, S. N.; Chasiotis, I. Molecular orientation and mechanical
property size effects in electrospun polyacrylonitrile nanofibers. Polymer 2011, 52,
1612–1618.
(344) Baji, A.; Mai, Y.-W.; Wong, S.-C.; Abtahi, M.; Chen, P. Electrospinning of polymer
nanofibers: Effects on oriented morphology, structures and tensile properties.
Composites Science and Technology 2010, 70, 703–718.
(345) Yarin, a. L.; Zussman, E. Upward needleless electrospinning of multiple nanofibers.
Polymer 2004, 45, 2977–2980.
(346) Lukas, D.; Sarkar, A.; Pokorny, P. Self-organization of jets in electrospinning from
free liquid surface: A generalized approach. Journal of Applied Physics 2008, 103,
084309.
(347) Miloh, T.; Spivak, B.; Yarin, a. L. Needleless electrospinning: Electrically driven
instability and multiple jetting from the free surface of a spherical liquid layer. Journal
of Applied Physics 2009, 106, 114910.
(348) Forward, K. M.; Rutledge, G. C. Free surface electrospinning from a wire electrode.
Chemical Engineering Journal 2012, 183, 492–503.
149
(349) Goring, D. A. I. The Lignin Paradigm. In Lignin: Properties and Materials; Glasser,
W.G.; Sarkanen, S., Ed.; ACS Symposium Series 397, American Chemical Society,
1989; pp. 2–10.
(350) Zhang, L. L.; Zhao, X. S. Carbon-based materials as supercapacitor electrodes.
Chemical Society reviews 2009, 38, 2520–31.
150
Appendix:
Viscosity (mPa*s)
concentration
SKL run
blend run
(wt%)
log(conc.)
1
run 2
1
10
1.000
2.3
2.3
2.6
15
1.176
3.7
3.7
4.5
20
1.301
6.7
6.7
7.7
25
1.398
12.1
12
15.9
30
1.477
24.6
25
38.5
35
1.544
56.4
55
83.2
37
1.568
83.3
84 147.9
39
1.591
126.3 127 204.3
40
1.602
172.4 184 291.3
43
1.633
327.4 356 472.4
45
1.653
585.1 536 760.2
blend
run 2
2.6
4.5
7.7
15.9
38.5
88.9
138.8
229.5
294.9
410.2
672.4
Specific Viscosity (h-hsolvent)/hsolvent (unitless)
SKL
blend blend run
run 1 run 2 Avg run 1
2
Avg
1.9
1.9
1.9
2.3
2.3
2.3
3.6
3.6
3.6
4.6
4.6
4.6
7.4
7.4
7.4
8.6
8.6
8.6
14.1 14.1 14.1
18.9
18.9
18.9
29.8 29.8 29.8
47.1
47.1
47.1
69.5 67.1 68.3 103.0 110.1 106.6
103.1 104.0 103.6 183.9 172.5 178.2
156.9 157.5 157.2 254.4 285.9 270.1
214.5 228.9 221.7 363.1 367.6 365.4
408.3 443.9 426.1 589.5 511.8 550.6
730.4 669.3 699.8 949.3 839.5 894.4
Log(specific viscosity)
SKL
blend
run 1 run 2 run 1
0.273 0.273 0.352
0.559 0.559 0.665
0.868 0.868 0.936
1.150 1.150 1.276
1.473 1.473 1.673
1.842 1.827 2.013
2.013 2.017 2.265
2.196 2.197 2.405
2.331 2.360 2.560
2.611 2.647 2.770
2.864 2.826 2.977
blend
run 2
0.352
0.665
0.936
1.276
1.673
2.042
2.237
2.456
2.565
2.709
2.924
Figure A1: Raw data for plot of specific viscosity vs. concentration (Figure 3.5)
151
Figure A2: Fitting to obtain slopes of specific viscosity vs. concentration (Figure 3.5)
152
Shear
Stress
(Pa)
F4, 40 wt%
G'
(Pa)
1.00
0.071
1.26
0.074
1.59
0.074
2.00
0.077
2.51
0.074
3.16
0.072
3.98
0.070
5.01
0.066
6.31
0.057
7.94
0.047
10.00 0.042
12.59 0.037
15.85 0.033
19.95 0.031
25.12 0.028
31.62 0.025
39.81 0.023
50.12 0.021
63.10 0.019
79.43 0.015
100.00 0.008
125.90 0.002
158.50 0.0001
199.50
251.20
316.20
398.10
501.20
631.00
794.30
1000.00
G"
(Pa)
4.752
4.759
4.761
4.763
4.768
4.769
4.774
4.779
4.788
4.799
4.804
4.806
4.807
4.808
4.809
4.809
4.815
4.816
4.818
4.821
4.829
4.833
4.835
4.833
4.835
4.836
4.836
4.841
4.845
4.851
4.859
F4, 40 wt%
PEO1M, 0.1
wt%
F4, 40 wt%
PEO1M, 0.2
wt%
F4, 40 wt%
PEO5M, 0.1
wt%
F4, 40 wt%
PEO5M, 0.2
wt%
G'
(Pa)
0.300
0.319
0.317
0.316
0.314
0.308
0.300
0.290
0.269
0.240
0.215
0.186
0.154
0.129
0.108
0.090
0.077
0.066
0.055
0.042
0.028
0.015
0.005
G'
(Pa)
0.303
0.307
0.305
0.308
0.304
0.300
0.296
0.290
0.276
0.259
0.243
0.227
0.209
0.191
0.172
0.152
0.135
0.119
0.105
0.089
0.073
0.058
0.048
0.039
0.032
0.025
0.018
0.010
0.002
G'
(Pa)
0.386
0.387
0.386
0.384
0.376
0.370
0.362
0.353
0.334
0.318
0.300
0.281
0.258
0.237
0.213
0.189
0.169
0.153
0.137
0.120
0.102
0.086
0.074
0.063
0.054
0.046
0.039
0.036
0.036
0.030
0.018
G'
(Pa)
0.590
0.561
0.571
0.561
0.546
0.546
0.535
0.523
0.508
0.489
0.470
0.443
0.417
0.387
0.356
0.322
0.291
0.262
0.235
0.209
0.184
0.163
0.147
0.134
0.123
0.110
0.101
0.103
0.095
0.065
0.022
G"
(Pa)
5.445
5.447
5.446
5.447
5.449
5.452
5.454
5.459
5.470
5.481
5.483
5.480
5.473
5.462
5.447
5.432
5.419
5.403
5.387
5.371
5.357
5.333
5.308
5.284
5.265
5.241
5.218
5.196
5.178
5.167
5.152
G"
(Pa)
5.344
5.339
5.343
5.346
5.350
5.355
5.359
5.361
5.372
5.378
5.379
5.376
5.370
5.360
5.346
5.330
5.308
5.284
5.257
5.229
5.200
5.169
5.135
5.097
5.060
5.023
4.983
4.947
4.917
4.899
4.895
G"
(Pa)
5.371
5.370
5.369
5.371
5.372
5.372
5.374
5.373
5.381
5.383
5.382
5.377
5.367
5.356
5.338
5.320
5.298
5.274
5.248
5.224
5.201
5.174
5.146
5.116
5.085
5.055
5.030
5.017
5.021
5.037
5.043
G"
(Pa)
5.484
5.481
5.488
5.492
5.488
5.493
5.495
5.499
5.504
5.511
5.513
5.505
5.497
5.482
5.460
5.433
5.400
5.366
5.324
5.285
5.245
5.202
5.158
5.110
5.065
5.026
5.001
4.993
4.987
4.971
4.946
Figure A3: Raw data for stress sweep (Figure 4.1a)
153
F4, 40 wt%
F4 40 wt%
PEO1M 0.1 wt%
F4 40 wt%
PEO1M 0.2 wt%
F4 40 wt%
PEO5M 0.1 wt%
F4 40 wt%
PEO5M 0.2 wt%
151G 40% 0.2% 5M
frequency
(rad/s)
1.00
1.26
1.59
2.00
2.51
3.16
3.98
5.01
6.31
7.94
10.00
12.59
15.84
19.95
25.12
31.63
39.81
50.11
63.09
79.43
100.00
125.90
158.50
199.50
251.20
316.20
398.10
501.20
627.80
G'
(Pa)
0.00
0.00
0.00
0.00
0.01
0.01
0.02
0.03
0.04
0.06
0.09
0.13
0.18
0.30
0.37
0.57
0.03
complex
complex
complex
complex
complex
G" viscosity G'
G'' viscosity
G'
G'' viscosity G'
G''
viscosity
G'
G'' viscosity
(Pa) (Pa*s) (Pa) (Pa)
(Pa*s)
(Pa) (Pa) (Pa*s) (Pa) (Pa)
(Pa*s)
(Pa) (Pa) (Pa*s)
0.51
0.51
0.00 0.57
0.57
0.00 0.57
0.57
0.01 0.57
0.57
0.02 0.62
0.62
0.65
0.52
0.01 0.72
0.57
0.01 0.72
0.57
0.01 0.71
0.57
0.02 0.77
0.61
0.82
0.52
0.01 0.90
0.57
0.01 0.90
0.57
0.02 0.90
0.57
0.03 0.97
0.61
1.03
0.52
0.01 1.14
0.57
0.02 1.14
0.57
0.02 1.13
0.56
0.05 1.22
0.61
1.30
0.52
0.02 1.43
0.57
0.03 1.43
0.57
0.04 1.42
0.56
0.08 1.53
0.61
1.64
0.52
0.03 1.80
0.57
0.05 1.80
0.57
0.05 1.78
0.56
0.11 1.91
0.60
2.07
0.52
0.05 2.26
0.57
0.07 2.25
0.57
0.08 2.23
0.56
0.16 2.38
0.60
2.60
0.52
0.07 2.84
0.57
0.11 2.83
0.56
0.12 2.79
0.56
0.23 2.97
0.59
3.28
0.52
0.10 3.57
0.57
0.15 3.55
0.56
0.17 3.50
0.56
0.31 3.69
0.59
4.13
0.52
0.14 4.49
0.56
0.22 4.44
0.56
0.23 4.38
0.55
0.42 4.58
0.58
5.19
0.52
0.19 5.63
0.56
0.30 5.57
0.56
0.33 5.47
0.55
0.56 5.69
0.57
6.54
0.52
0.26 7.05
0.56
0.42 6.96
0.55
0.44 6.83
0.54
0.75 7.06
0.56
8.24
0.52
0.36 8.85
0.56
0.60 8.69
0.55
0.60 8.53
0.54
0.97 8.75
0.56
10.36
0.52
0.51 11.08
0.56
0.79 10.87
0.55
0.82 10.64
0.53
1.32 10.83
0.55
13.08
0.52
0.54 13.95
0.56
1.29 13.48
0.54
1.19 13.23
0.53
1.42 13.57
0.54
16.51
0.52
0.84 17.42
0.55
1.37 16.97
0.54
1.42 16.57
0.53
2.27 16.64
0.53
21.02
0.53
0.18 22.09
0.55
1.42 21.28
0.54
1.35 20.82
0.52
4.63 20.18
0.52
26.60
0.53
27.83
0.56
2.19 26.45
0.53
2.16 25.93
0.52
7.87 24.71
0.52
33.74
0.53
34.28
0.54
10.90 31.61
0.53
4.25 32.17
0.51
11.42 30.84
0.52
43.31
0.55
43.37
0.55
26.26 38.14
0.58
10.46 39.84
0.52
15.66 38.53
0.52
54.52
0.55
54.58
0.55
23.48 48.93
0.54
17.91 49.93
0.53
27.31 49.05
0.56
70.20
0.56
70.66
0.56
3.38 65.13
0.52
25.78 63.29
0.54
46.57 56.93
0.58
84.78
0.53
89.08
0.56
83.01
0.52
41.87 80.52
0.57
76.72 75.03
0.68
109.20 0.55
111.60
0.56
108.1
0.54
18.88 102.70
0.52
63.09 99.19
0.59
148.70 0.59
132.40
0.53
138.4
0.55
146.30
0.58
38.90 133.5
0.55
192.70 0.61
211.00
0.67
120.2
0.38
191.60
0.61
29.15 164.9
0.53
244.10 0.61
206.60
0.52
212.5
0.53
203.10
0.51
213.0
0.54
367.60 0.73
225.30
0.45
303.8
0.61
236.30
0.47
239.9
0.48
383.70 0.61
165.60
0.26
160.1
0.26
188.20
0.30
306.3
0.49
Figure A4: Raw data for frequency sweep (Figure 4.2b)
154
Figure A5: Example of CaBER data to obtain relaxation times (Chapter 4). (a) Linear scale plot of D mid(t)/D1
vs. time. (b) Semi-log plot of data depicted in (a). (c) Fitting of an exponential decay to the region
corresponding to elastocapillary thinning (t = 0.02- 0.1 s), on the semi-log plot shown in (b). Using the fitting
data and equation (1) from Section 2.6.3: 13.32 = (1/3)   = 0.025 s.
155
Figure A6: Aliphatic region from HSQC of F4SKL in DMSO-d6. Peaks in this region were not assigned to
specific linkage structures.
156
Figure A7: Aromatic region from HSQC of F4SKL in DMSO-d6. Peaks in this region were not assigned to
specific linkage structures.
157
Figure A8: Aliphatic region from HSQC of F1-3SKL in DMSO-d6. Peaks in this region were not assigned to
specific linkage structures.
158
Figure A9: Aromatic region from HSQC of F1-3SKL in DMSO-d6. Peaks in this region were not assigned to
specific linkage structures.
159
Figure A10: Wide angle X-ray diffraction patterns of BF (top) and NBF (bottom) carbonized at different
temperatures, 600 (black), 800, (red), and 1000 oC (blue).
160