Miscibility and hydrogen bonding in blends of poly(ethylene oxide) and kraft lignin

Article abstract of DOI:10.1021/ma0348371

Polymer blending is a convenient method to develop products with desirable properties. Through specific intermolecular interactions favorable polymer blending can occur, and composite materials with desirable properties can be produced. In this study we h

Full text of DOI:10.1021/ma0348371

Macromolecules 2003, 36, 7803-7811
7803
Miscibility and Hydrogen Bonding in Blends of Poly(ethylene oxide) and
Kraft Lignin
J oh n F . Ka d la * a n d Sa tosh i Ku bo
College of Natural Resources, North Carolina State University, Raleigh, North Carolina 27695-8005
Received J une 19, 2003; Revised Manuscript Received J uly 15, 2003
ABSTRACT: Polymer blending is a convenient method to develop products with desirable properties.
Through specific intermolecular interactions favorable polymer blending can occur, and composite
materials with desirable properties can be produced. In this study we have prepared poly(ethylene oxide)-
lignin blends using thermal mixing. Miscible blends were observed over the entire blend ratio. A melting
point depression, comparable to results obtained from phenoxy/PEO blends, and a negative deviation of
Tg from the weight-average values was observed. The effect of the binary interaction parameter, ø, on Tg
was analyzed. Satisfactory prediction of the Tg-composition curve was obtained, in which specific
intermolecular interactions exist. FT-IR analyses revealed a strong hydrogen bond between the aromatic
hydroxyl proton and the ether oxygen in PEO.
In tr od u ction
namic mechanical analysis,13 microscopy,14 and solid-
state NMR.15 Thermal analysis, specifically DSC can
determine whether two or more polymers are miscible
or immiscible. A single compositional-dependent glass
transition temperature (Tg) is an indication of full
miscibility at a dimensional scale between 5 and 15
nm.11 On the basis of the composition of the blend, the
Tgs of the individual components, and the nature of the
interactions between these components, several theo-
retical and empirical equations have been proposed for
the prediction of blend Tg.16-23
Recently, there has been a rising interest in bioma-
terials utilizing natural polymers, owing to the low cost,
biodegradability, and biocompatibility.1 Polymer blend-
ing is of considerable importance as an alternative to
graft copolymerization and the high cost to develop
new homopolymers. Extensive research has been un-
dertaken in blending different polymers to obtain new
products having some of the desired properties of each
component.2-5
Lignin, a natural polymer found in wood, is a pol-
yaromatic polyol that is readily available and relatively
inexpensive. Lignin structure is dependent on wood
species and processing conditions.6 Commercially, lignin
is obtained as a byproduct of “wood free” papermaking.
Some industrial processes consistently produce standard
well-defined lignins. As a result, lignin is utilized as a
stabilizer (antioxidant) for plastics and rubber as well
as in the formulation of dispersants, adhesives and
surfactants. Lignin has also been used in the production
of fully biodegradable lignin grafted polyolefins.7
Extensive cross-linking and strong intramolecular
interactions of polymeric lignins constrain the utiliza-
tion of polymeric lignin in solid material systems.6,8
Through polymer blending these interactions can be
disrupted, thus altering the lignin’s viscoelastic proper-
ties.6 However, as with most polymeric systems, the
observation of miscible polymer blends with lignin are
rare.9
A powerful tool to detect specific intermolecular
interactions in polymer blends is Fourier transform
infrared (FT-IR) spectroscopy. Observed changes in
hydroxyl, carbonyl and ether vibrations provide direct
evidence of specific interactions between blend compo-
nents.12,24 The magnitude of the shift in wavenumber
(∆ν) arising due to blend formation yields a measure of
the average strength of the intermolecular interac-
tions.25
In this paper, the miscibility of lignin/poly(ethylene
oxide) (PEO) blends of varying composition was mea-
sured by thermal analysis. The interaction parameter
and energy density between blend components were
estimated using melting point depression and glass
transition temperature measurements. FT-IR spectros-
copy was used to study the intermolecular interactions
between lignin and PEO. Using lignin model com-
pounds, the role of the aliphatic and phenolic hydroxyl
groups in the interaction with PEO was determined.
Polymer blending is a convenient method to develop
products with desirable properties. The chemical and
physical properties of the polymer blends are dependent
on monomer type(s), molecular weight, and distribution
and composition of the respective polymers. Most poly-
mers however are immiscible due to the low entropy of
mixing.10 Only through specific intermolecular interac-
tions can favorable polymer blending occur and com-
posite materials with desirable properties be produced.
Exp er im en ta l Section
Ma ter ia ls. Hardwood kraft lignin was obtained from West-
vaco Corp. The lignin was repeatedly washed with dilute HCl
(Aldrich Chemicals) to exchange sodium counterions, vacuum-
dried over P2O5, and recovered as a fine powder. Chemical
properties of the lignin are listed in Table 1. Poly(ethylene
oxide) (PEO), 100 000 amu (Union Carbide Corp.) and poly-
styrene (PS), 140 000 amu (Aldrich Chemicals) were used as
received. Sodium borohydride (NaBH4), sodium hydroxide
(NaOH), and all solvents were purchased from Aldrich Chemi-
cals and used as received. Carbon tetrachloride used for
FT-IR measurement was dried over 4A activated molecular
sieves before use. 3,4-Dimethoxy acetophenone, 2-methoxy-4-
hydroxyphenol, 3-(3,4-dimethoxy phenyl)-1-propanol (I), 2-meth-
Various techniques exist to determine the miscibility
of polymer blends,11 including thermal analysis,12 dy-
* Corresponding author. Telephone: (919) 513-2455. Fax: (919)
515-6302. E-mail: jfkadla@ncsu.edu.
10.1021/ma0348371 CCC: $25.00 © 2003 American Chemical Society
Published on Web 09/10/2003

7804 Kadla and Kubo
Macromolecules, Vol. 36, No. 20, 2003
Ta ble 1. Sp in n in g Tem p er a tu r e of Lign in /P EO Blen d s
Diazomethane methylation was repeated. After the final
methylation, the solution was removed by evaporation and the
residue was washed with diethyl ether (500 mL) and dried.
The completeness of methylation was determined using the
∆ꢀ method27 and the absence of a bathochromic shift in the
UV difference spectra. In addition, 1H NMR analysis of ace-
tylated methylated lignin showed no phenolic acetate groups.28
Blen d P r ep a r a tion . Prior to mechanical blending the
lignin was thermally treated at 145 °C under vacuum for 30
min to remove volatile contaminants.29 Blends of various PEO/
lignin ratios were prepared by mechanical mixing followed by
thermal extrusion using an Atlas Laboratory Mixing Extruder
(Atlas Corp.). Extrusion temperatures were varied between
130 and 240 °C depending on the blend composition (Table
1).
blend ratio (weight)
spinning temp (°C)
lignin
PEO
lignin/PEO
M-lignin/PEO
100
95
0
5
208-215
205
166-170
87.5
75
62.5
60
50
40
25
10
0
12.5
25
37.5
40
50
60
75
90
100
187-190
168-170
162-170
161-168
168-169
168-174
153-158
204-210
229-230
230-231
216-222
165
164-177
210
220-232
216-222
The lignin model compound/PEO and PS blends were
prepared by mixing 100 mg of polymer with 0.5 mL of a 0.2
mol L-1 lignin model compound in CHCl3 (1 mmol of lignin
model compound dissolved in 5 mL of CHCl3). After complete
dissolution, the solvent (CHCl3) was removed under reduced
pressure. The blend (100 mg) was then set between glass plates
and heated at 210 °C for 5 min under a N2 atmosphere. The
sample was then cooled to room temperature and dried in a
vacuum desiccator.
Ch a r a cter iza tion of Lign in /P EO Blen d s. Mechanical
properties of lignin/PEO blend fibers were estimated as tensile
properties using an Instron model 4411 equipped with a 0.5
N load cell. The test speed was 0.5 mm min-1, and the sample
length was 25 mm. Tensile tests were performed at constant
temperature (25 °C) and relative humidity (65%). All reported
results are the average of 20 tests.
Glass transition temperatures, Tg, of the blends were
determined on a TA Instruments Q 100 DSC with a scan rate
of 20 °C/min over the temperature range of -90 to 200 °C.
The measurements were made using 4-5 mg samples under
a nitrogen atmosphere after the samples were quickly cooled
to -90 °C. The glass transition temperature was recorded as
the midpoint temperature of the heat capacity transition of
the second heating run. Samples were run in duplicate and
are reported as the average of the two runs and were within
experimental error of each other ((1.0 °C).
oxyphenol (III), and 2,6-dimethoxyphenol (IV) and phenol,
were purchased from Aldrich Chemicals and recrystallized
from benzene/petroleum ether (6/4).
1-(3,4-Dim eth oxyp h en yl)eth a n ol (II) was prepared by
reacting 2 equiv of NaBH4 (0.42 g, 11.0 mmol) with 3,4-
dimethoxy acetophenone (1.0 g, 5.5 mmol) in 3:1 EtOH:H2O
(50 mL) and heated under reflux for 3 h. The reaction mixture
was then cooled, neutralized by bubbling CO2 through the
supernatant solution, and extracted with 1,2-dichloroethane
(3 × 50 mL). A quantitative conversion of the acetophenone
was obtained. MS m/z (relative intensity) 182 (M+, 59), 167
(87), 153 (47), 139 (100), 124 (32), 108 (21), 93 (50), 77 (21), 65
1
(25), 51 (11), 43 (41). H NMR δ (ppm): 1.48 (d, 3H), 3.90 (d,
6H), 4.83 (q, 1H), 6.84 (q, 1H), 6.86 (q, 1H), 6.93 (d, 1H).
3,3′-Dim eth oxy-5,5′-dim eth yl-[1,1′-biph en yl]-2,2′-diol (V)
was prepared by reacting a solution of 0.25 g (2.0 mmol) of III
in a mixture of 5.0 mL of EtOH and 1.0 mL of 10% aqueous
NaOH at 5 °C with a solution of 0.82 g (2.5 mmol) of K3Fe-
(CN)6 in 5 mL of water followed by 5 mL of EtOH and 1.0 mL
of 10% aqueous NaOH to facilitate stirring. After 2.0 h, the
reaction mixture was diluted with 50 mL of saturated NH4Cl,
adjusted to pH 5.5, and extracted with EtOAc, and the dried
(MgSO4) extract evaporated in vacuo to leave 0.21 g of crude
product. Recrystallization from benzene-petroleum ether af-
forded colorless crystals, Tm ) 125 °C (DSC) 1H NMR δ
(ppm): 2.30 (s, 6H); 3.84 (s, 6H); 6.04 (s, D2O-exchangeable),
6.68 (s, 2H), 6.72 (s, 2H).
The equilibrium melting point, Tmeq, was determined by
DSC using Hoffman-Weeks plots.30 In a typical experiment,
5.0 mg samples, as weight fractions of PEO, were heated to
90 °C and maintained at this temperature for 10 min to
completely eliminate PEO crystallinity. The sample was than
quenched to the desired isothermal crystallization tempera-
ture, Tic, and held at temperature for 2 h to allow complete
crystallization. After isothermal crystallization, the melting
temperature, Tm′, was measured using a heating rate of 10
°C min-1; Tm′ was determined as the peak top temperature.
Infrared (FT-IR) spectra of the polymer blends were deter-
mined using the diffuse reflectance (DRFT-IR) method (due
to the physical properties of the polymer blends it was difficult
to prepare uniform KBr pellets for FT-IR analysis using
transmittance detection). The polymer blends (10 mg) were
dispersed in KBr (200 mg), and DRFT-IR measurements were
recorded on a Perkin-Elmer 16PG FT-IR spectrometer; 256
Lign in Ch a r a cter iza tion . Density values were deter-
mined according to the ASTM standard (D70-97) using a
multivolume pycnometer 1305 (Micromeritics). Elemental
analysis of the lignin samples was carried out at E & R
Microanalytical Laboratories Inc. The methoxyl content was
determined at our laboratory according to the modified pro-
cedure of Viebock and Schwappach.26 Aliphatic and aromatic
1
hydroxyl contents were determined using H NMR. Quantifi-
cation was obtained from the integration ratios of aliphatic
and aromatic acetoxy protons of acetylated lignin preparations
relative to the internal standard, p-nitrobenzaldehyde.
The lignin preparations were acetylated by dissolving 200
mg of the lignin in 10 mL of pyridine-acetic anhydride (1:1,
v/v) and reacted for 48 h at room temperature. The solution
was poured over crushed ice and filtered. The resulting
precipitate was then washed with cold water/HCl, dried, and
subjected to a second acetylation treatment.
scans were collected with a spectral resolution of 2.0 cm-1
.
Owing to the hygroscopic nature of the polymer blends, a pure
nitrogen flow was maintained over the sample during collec-
tion. FT-IR analysis of the model compounds was performed
in CCl4 using a liquid cell with ZnSe windows and a 1 mm
path length, 16-32 scans were acquired at a spectral resolu-
The average molecular mass and molecular mass distribu-
tion of acetylated lignin samples were determined by GPC
(Waters Associates, UV and RI detectors) using styragel
columns at 50 °C and THF as the eluting solvent. The GPC
system was calibrated by using standard polystyrene samples.
The injection volume was 100 µL, and the acetylated lignin
concentration was 1 mg mL-1 THF.
Meth yla tion of Lign in . Methylation of the phenolic hy-
droxyl groups in lignin was performed using a diazomethane
ether solution produced from N-methyl-N-nitroso-p-toluene-
sulfonamide. Accordingly, 10.0 g of lignin was suspended in
dioxane/methanol (200 mL; 2:1, v/v) followed by the addition
of the diazomethane ether solution (50 mL) and stirred for 3
h at room temperature. The ether phase was removed by
evaporation and fresh diazomethane (20 mL) was added.
tion of between 2.0 and 4.0 cm-1
.
Resu lts a n d Discu ssion
P r ep a r a tion of Lign in /P EO Blen d F iber s. Extru-
sion temperatures of the lignin/PEO blends are listed
in Table 1. All samples exhibited good thermal process-
ing properties. Extrusion temperature was dependent
on blend composition, with the lowest temperature
observed in the middle range of the blending ratio.

Macromolecules, Vol. 36, No. 20, 2003
Poly(ethylene oxide)-Lignin Blends 7805
F igu r e 2. DSC curves of lignin/PEO (A) and M-lignin/PEO
(B) blend fibers.
Ta ble 3. Ch em ica l P r op er ties of Ha r d w ood Kr a ft Lign in
functional goups (mmol g-1
hydroxyl
)
molecular mass
aliphatic
4.1
phenolic
4.3
methoxyl
5.8
S G-1
1.2
Mw
dispersity
1.8
F igu r e 1. Tensile strength and Young’s modulus of lignin/
PEO blend fibers.
2400
interactions between components: consistent with a
miscible polymer blend system.32
Ta ble 2. P h ysica l P r op er ties of Lign in /P EO Blen d F iber s
blend ratio
(weight)
tensile
strength
(MPa)
Young’s
modulus
(GPa)
An a lysis of In t er m olecu la r In t er a ct ion s b e-
tw een P EO a n d Lign in . Lignin, a polyol, has both
phenolic and aliphatic hydroxyl groups (Table 3) that
can participate in hydrogen-bonding with PEO (vide
infra). To investigate the role of the various hydroxyl
groups and to characterize the hydrogen-bonding inter-
actions between lignin and PEO, the phenolic hydroxyl
groups in lignin were selectively methylated using
diazomethane.33 The methylated lignin (M-lignin) was
then blended with PEO under the same conditions as
those used in the unmethylated lignin blends.
diameter
(µm)
elongation
(%)
lignin PEO
100
95
87.5
75
50
25
0
0
5
148 ( 25 23.1 ( 3.3 3.85 ( 0.27
162 ( 11 24.7 ( 2.5 4.19 ( 0.31
0.6 ( 0.1
0.6 ( 0.1
0.6 ( 0.1
0.8 ( 0.1
3.5 ( 0.5
12.5 163 ( 21 22.5 ( 1.6 4.04 ( 0.20
25
50
75
100
151 ( 7
185 ( 27 15.1 ( 0.7 1.07 ( 0.08
114 ( 5
274 ( 16 12.6 ( 0.2 0.51 ( 0.03
16.4 ( 1.5 2.35 ( 0.08
13.2 ( 3.3 0.98 ( 0.04 19.7 ( 3.7
5.4 ( 0.6
The Young’s modulus (3.9 GPa), tensile strength (23
MPa) and elongation (0.6%) of the lignin-based blends
increased slightly with addition of low wt % additions
of PEO (Table 2). Incorporation of 5 wt % of PEO
increased the corresponding values to 4.2 GPa, 25 MPa,
and 0.6%, respectively (Figure 1). Increasing addition
of PEO reduced both the Young’s modulus and tensile
strength of the lignin-based thermoplastics, decreasing
dramatically at PEO levels of 25-50 wt %. The initial
increase in tensile behavior is likely the result of PEO
disrupting the intermolecular association of the lignin
macromolecules; the physical properties of lignin-based
materials are determined by the extent of lignin inter-
molecular association and formation of supramacromo-
lecular complexes.31 Through specific intermolecular
interactions between PEO and individual lignin com-
ponents, the supramacromolecular lignin complexes
dismantle. Increasing PEO incorporation further dis-
rupts the supramacromolecular lignin structure and the
observed physical properties become more strongly
influenced by the PEO component. At 75 wt % PEO,
the lignin/PEO blend exhibits a substantial increase in
elongation, 19.7%, compared with 0.6% and 5.4% for
lignin and PEO, respectively. This large change in
ductility suggests the elimination of the supramacro-
molecular lignin complexes and the formation of specific
Th er m a l An a lysis. Differential scanning calorimetry
(DSC) is extensively used to investigate miscibility in
polymer blends. A single compositional-dependent glass
transition temperature (Tg) is an indication of full misc-
ibility at a dimensional scale between 5 and 15 nm.11
Figure 2 shows DSC analyses of the lignin/PEO and
M-lignin/PEO blends of various compositions. The pure
amorphous lignins exhibit one Tg at 106 and 86 °C, re-
spectively. In both systems, the Tg decreases in tem-
perature with increasing PEO blend content. Similarly,
the melting temperature (Tm) of the PEO component in
the blends decreases with the increase of lignin content.
Table 4 summarizes the thermal properties of PEO,
lignin, M-lignin, and the corresponding polymer blends.

7806 Kadla and Kubo
Macromolecules, Vol. 36, No. 20, 2003
Ta ble 4. DSC Resu lts of th e Lign in /P EO a n d M-Lign in /P EO Blen d F iber s
lignin/PEO blend M-lignin/PEO blend
blend ratio (weight)
Tg
(°C)
∆Cp
Tm
(°C)
∆H
crystallinity
(%)
Tg
(°C)
∆Cp
Tm
(°C)
∆H
crystallinity
(%)
lignin
PEO
(J g-1 C-1
)
(J g-1
)
(J g-1 C-1
)
(J g-1
)
1
0
106
95
76
38
5
0.60
0.61
0.68
0.77
0.85
a
0.73
0.49b
0.26b
0.18
0.11
×
×
0
0
0
0
0
a
1
63
112
146
168
0
0
0
0
0
a
1
56
80
86
90
86
a
48
13
a
0.53
a
0.71
0.75
a
0.62
0.35
0.24b
0.25b
0.12
0.11
×
0
a
0
0
a
0
a
0
0
a
0.95
0.875
0.75
0.625
0.6
0.5
0.4
0.25
0.1
0
0.05
0.125
0.25
0.375
0.4
0.5
0.6
0.75
0.9
1
a
×
×
×
×
×
a
a
a
-15
-7
4b
54
56b
57
61
63
67
2
3
-15
-7b
4b
-2.4
-50
53
56
61
66
67
56b
73
113
145
168
60b
65
80
86
90
-5b
-35
-50
a
b
No sample. There are some difficulties in assigning the thermal transition point because the melting or recrystallization peak is
partially overlapped with the glass-transition state and the melting peak.
The detection of a single Tg via thermal analysis
provides important information as to the composition
of the individual amorphous phases present in the
material. All lignin/PEO and M-lignin/PEO blends show
only a single Tg over the entire blend composition. The
single Tg strongly suggests that these are fully miscible
blends with a homogeneous amorphous phase. In gen-
eral, the formation of miscible binary polymer blends
depends on the occurrence of exothermic interactions
between the two components being mixed.10 The manner
in which the glass transition temperature varies with
blend composition is reflective of the relative strengths
of these intermolecular interactions. Several theoretical
and empirical equations have been proposed for the
prediction of Tg in miscible polymer blends, based on
the composition of the blend, the Tgs of the individual
components, and the nature of the interactions between
these components.16,20,21,23 Of these, Lu and Weiss21
derived a general expression for the Tg of a binary
polymer mixture (Tgm) wherein the other popularly used
equations were shown to be special cases within their
theory. The equation included no adjustable parameters
and utilized entropy as the thermodynamic variable.
The enthalpy of mixing for the blend (∆Hm) was related
to the Flory-Huggins interaction parameter (ø) through
the van Laar relationship ∆Hm ) øRTφ1φ2, where φI is
the volume fraction of component i, R is the gas constant
and T the temperature. The Flory-Huggins interaction
parameter (ø) was assumed to be a quadratic function
polymer i), wi is the weight fractions of polymer i, and
∆cpi ) clpi - cgpi at Tgi (the change in specific heat of
polymer i), superscripts g and l denoting the glassy and
liquid states, respectively.
In polymer blends characterized by weak intermo-
lecular interactions between the miscible polymer blend
components, the effect of k is greater than A,21 and the
equation is simplified to the Gordon-Taylor equation:
20
(w1Tg1 + k′w2Tg2)
Tgm
)
(w1 + k′w2)
Here
k′ ) k + A/(Tg2 - 1)
and k′ can be employed qualitatively as a measure of
the intermolecular interactions in the polymer blend.21
By contrast, in polymer blends with strong intermo-
lecular interactions, the magnitudes of k, b, and c are
such that Tgm may be quantified by Kwei’s equation:22
(w1Tg1 + kw2Tg2)
Tgm
)
+ A′w1w2
(w1 + kw2)
where
2
of composition; ø ) øo + ø1w2 + ø2w2 where w2 is the
∆cp2
weight fraction of polymer 2. Similarly, the isobaric
specific heat of the mixture (cpm) was assumed to have
k )
∆cp1
the form cpm x1cp1 + x2cp2 + x1x2δcp, where xi is the
)
mole-fraction of polymer i, cpi the isobaric specific heat
of component i, and δcp is the change in specific heat
due to mixing (which is negative for a miscible blend).34
The resulting equations are21
In this case, A′ may be regarded as measure of the
interactions between polymer components.
The Lu and Weiss equation satisfactorily predicts Tg-
composition curves for miscible polymer blends that
exhibit either positive or negative deviations from a
linear mixing rule, depending on the strength of inter-
action.21 Figure 3 shows the Tg-composition curves for
the lignin/PEO and M-lignin/PEO blends. Both polymer
blend systems exhibit negative deviation from linear
mixing and is satisfactorily predicted by the Lu and
Weiss equation (solid line in Figure 3) as well as the
Kwei equation (dotted line in Figure 3). The Gordon-
Taylor equation did not produce a satisfactory fit of the
experimental data (dashed line in Figure 3). Table 5 lists
the fitting parameters obtained from the various theo-
retical and empirical equations used to predict Tg of
miscible polymer blends.
(w1Tg1 + kw2Tg2)
Tgm
)
+
(w1 + kw2)
Aw1w2
(w1 + kw2)(w1 + bw2)(w1 + cw2)2
- øR(Tg1 - Tg2)c
M1∆cp1
∆cp2 - w1δclp
∆cp1 - w2δcgp
A )
k )
where b ) M2/M1, (Mi is the molar mass per chain
segment of polymer i), c ) F1/F2 (Fi is the density of

Macromolecules, Vol. 36, No. 20, 2003
Poly(ethylene oxide)-Lignin Blends 7807
of a crystalline polymer blended with an amorphous
polymer provides important information about its mis-
cibility and its associated polymer-polymer interaction
parameter (ø). The temperature reduction is caused by
thermodynamic depression arising from a reduction in
chemical potential due to the presence of the polymeric
solvent. When two polymers are miscible in the molten
state, the chemical potential of the crystallizable poly-
mer is decreased due to the addition of the second
component. This leads to a reduction in the equilibrium
melting temperature with increasing amorphous poly-
mer content, especially in blends containing specific
interactions between components. An immiscible blend
will typically not show the depression of the equilibrium
melting point. Using the equation of Nishi and Wang,37
which is based on the Flory-Huggins theory,10 the
melting point depression is given by the following
equation:10,37
F igu r e 3. Tg of lignin/PEO (A) and M-lignin/PEO (B) blend
fibers: solid line, Lu and Weiss; dotted line, Kwei; dashed line,
Gordon-Taylor.
The calculated values of ø from the Lu and Weiss
equation (-0.34 and -0.15 for the lignin and M-lignin
PEO blends, respectively) indicate relatively strong
intermolecular interactions, wherein the lignin, contain-
ing the phenolic hydroxyl groups, forms stronger inter-
molecular interactions with PEO than the M-lignin,
with only aliphatic hydroxyl groups available.
The observed Tg phenomenon, which is related to
segmental motions in the polymer, is affected by steric
hindrance and variations of the flexibility of the polymer
chains.23,35 As a result, A′ has been related to both
hydrogen bond formation and to changes in the environ-
ment of the polymer chains. The methylation of the
phenolic hydroxyl groups within the lignin macromol-
ecule decreases the extent of intra and intermolecular
interactions within lignin, increasing the molecular
mobility or flexibility of the polymer chains and results
in the observed lower Tg of the M-lignin relative to the
original lignin. Therefore, the lower value of A′ (A′ )
-170 for the lignin/PEO cf. A′ ) -193 for M-lignin/PEO
blends) for the M-lignin/PEO blend indicates a lower
propensity to hydrogen-bonding than that of the lignin/
PEO blends.36
V2u
2
∆Tm ) T0m - T0m2 ) - Tm0
Bφ1
(
)
∆H2u
Subscripts 1 and 2 represent the amorphous and the
crystalline polymers, respectively, T0m is the equilibri-
um melting point of the pure crystallizable component,
T0m2 is the equilibrium melting point of the blend, φ is
the volume fraction, Vu is the molar volume of the
repeating units, ∆Hu is the heat of fusion per mole of
repeat unit, and B refers to the interaction energy
density characteristic of the polymer pair, and is, in
practice, related to the thermodynamic interaction
parameter ø12 by B ) RT0 (ø12/V1u).
To determine the equilimbrium melting temperatures
for PEO (T0m) and the polymer blends (Tm0 2) of given
compositions, and to eliminate the morphological effect
from the melting point depression, the method of
Hoffman and Weeks was used.30 However, as discussed
and demonstrated by many groups,38,39 the errors as-
sociated with the Hoffman-Weeks approach could be
Equ ilibr iu m Meltin g Tem p er a tu r e a n d Meltin g
P oin t Dep r ession . The depression of the melting point
Ta ble 5. Equ a tion s for Tg Va r ia tion s w ith Blen d Com p osition of Lign in /P EO Blen d F iber s
parameter
M-lignin/PEO
author(s)
Fox
equation
lignin/PEO
w1
w2
1
)
+
-
-
Tg Tg1 Tg2
(R2 ) 0.927)
(R2 ) 0.807)
w1∆cp1 ln Tg1 + w2∆cp2 ln Tg2
w1∆cp1 + w2∆cp2
Couchman
Gordon-Taylor
Kwei
ln Tg )
-
-
(R2 ) 0.936)
(R2 ) 0.894)
w1Tg1 + kw2Tg2
w1 + kw2
Tg )
Tg )
k ) 0.37 ( 0.04
(R2 ) 0.982)
k ) 0.27 ( 0.03
(R2 ) 0.988)
w1Tg1 + kw2Tg2
w1 + kw2
+ Aw1w1
q ) -170 ( 8
k ) 1
q ) -193 ( 4
k ) 1
(R2 ) 0.996)
(R2 ) 0.999)
w1Tg1 + kw2Tg2
w1 + kw2
Aw1w2
(w1 + kw2)(w1 + bw2)(w1 + cw2)2
Tg )
+
Lu-Weiss
ø ) -0.34
k ) 1.9
ø ) -0.15
k ) 0.8
(R2 ) 0.997)
(R2 ) 0.999)
øR(Tg1 - Tg2)c
∆cp2 - w1δclp
A ) -
k )
∆cp1 - w2δcgp
M1∆cp1
a
Here Tg, Tg1, and Tg2 are glass transition temperature of the blend and homopolymers, respectively. w1 and w2 are the corresponding
weight fractions. ∆cp1 and ∆cp2 are the magnitudes of increase in heat capacity at Tg of the corresponding homopolymers. b and c in the
Lu and Weiss equation are ratios of density, F1/F2, and molar mass of repeating unit, M2/M1, of homopolymers.

7808 Kadla and Kubo
Macromolecules, Vol. 36, No. 20, 2003
F igu r e 4. Tm and ∆H values of lignin/PEO ) 25/75 blend as
a function of the crystallization time. Triangle and circle
symbols are samples crystallized at 45 and 35 °C, respectively.
F igu r e 6. Equilibrium melting point depression of PEO in
blends: (b) lignin blend; (O) M-lignin blend.
of the melting point depression, ∆Tm, vs the square of
2
the volume fraction of lignin, φ1 (Figure 6), produces a
B value of -5.5 cal cm-3 for the lignin/PEO blend and
-2.9 cal cm-3 for the M-lignin/PEO blend (∆H2u ) 1.980
kcal/mol and V2u ) 38.1 cm3/mol 40). The negative B
value is consistent with a miscible system, and is
comparable to results obtained from Phenoxy/PEO
blends (B ) -6.2 cal cm-3).41 However, the T0m depres-
sion plots for both the lignin/PEO and M-lignin/PEO
blends deviate from a linear line. This may suggest a
composition dependence of ø, and therefore, ø values
could not be calculated, for these systems. Nonetheless,
a melting point depression is clearly observed (Table 4)
for both polymer blend systems, suggesting a negative
value for ø40 and a stronger interaction between the
lignin/PEO blends than the M-lignin/PEO blends.
Fou r ier Tr an sfor m In fr ar ed Spectr oscopy An aly-
sis. Infrared spectroscopy has been proven to be a highly
effective means of investigating specific interactions
between polymers. FT-IR can be used to qualitatively
and quantitatively study the mechanism of interpolymer
miscibility through hydrogen-bonding. To further un-
derstand the thermodynamic behavior of the lignin-
based/PEO blends the intermolecular interactions of
PEO with lignin and M-lignin was investigated with FT-
IR.
Figure 7 shows the FT-IR spectra of the hydroxyl-
stretching (νO-H) region (3100-3700 cm-1) of lignin (A)
and M-lignin (B) and the respective PEO blends. Owing
to the variety of intra- and intermolecular hydrogen
bonds within and between lignin macromolecules a
broad νO-H band is observed for both lignins. The lignin
spectrum has a broad band center at ∼3422 cm-1 with
an apparent shoulder at ∼3518 cm-1. These bands are
assigned to the average stretching of intermolecular
hydrogen bonding and intramolecular hydrogen bond-
ing, respectively. The M-lignin exhibits the same broad
band center at ∼3422 cm-1, but the shoulder at ∼3518
cm-1 is less pronounced.
F igu r e 5. Hoffman-Weeks plot for isothermally crystallized
lignin/PEO (A) and M-lignin/PEO (B) blend fibers: (a) 0/100;
(b) 10/90; (c) 25/75; (d) 32.5/67.5; (e) 35/65; (f) 40/60; (g) 50-
50.
significant, and a small uncertainty in melting point
could profoundly affect the value of ø. In certain blends,
the appearance of a melting point depression may in
fact be the result of incomplete crystallization, rather
than thermodynamic considerations. This is particularly
the case in polymer blend systems with very small
melting point depressions. The effect of crystallization
time on the observed melting point was determined for
our blend systems and is illustrated in Figure 4. The
Tm of the PEO fraction in lignin/PEO ) 25/75 blend is
independent from crystallization period, if that period
is longer than 2 h. This behavior is observed for both
high and low crystallization temperatures, 45 and 35
°C, respectively. The difference in observed Tm between
crystallization times of 2 and 12 h was measured to be
less than 0.5 °C (within experimental error of the
instrument usedsDSC analysis). The observed ∆H
value at 2 h is >95% of relative ∆H value for a
crystallization time of 12 h. Therefore, all Tm measure-
ments were made using crystallization times of 2 h and
with relative ∆H values within 95% of the maximum
∆H obtained for that blend.
Blending with PEO dramatically changes the νO-H
region of the lignin. In the lignin/PEO blends (Figure
7A) the broad center band shifts to lower wavenumber
with increasing PEO content, and a third band center
appears at ∼3295 cm-1. The relative intensity of this
band increases with increasing PEO content, and the
shape of the νO-H region becomes broader. In contrast,
the broad center band in the M-lignin/PEO blends does
not shift, and less change in the overall νO-H region is
observed (Figure 7B).
Figure 5 presents the Hoffman-Weeks plot to obtain
the equilibrium melting points for pure PEO, the lignin/
PEO (A) and the M-lignin/PEO (B) blends. The experi-
mental results were fit using T′ ) σTic + (1 - σ)
m
Tm0 ,30,37 where σ is a stability parameter37 which is
usually related to morphological factors, and Tic is the
isothermal crystallization temperature. The values of
T0m for PEO and its blends with the lignin were ob-
tained by the extrapolation procedure using a least-
squares fit of the data to the intersection with T′ )
A corresponding effect is observed in the asymmetric
m
Tic. Assuming B or ø is composition independent, a plot
C-O-C stretching (νas(C-O-C)) region (1050-1150 cm-1
)

Macromolecules, Vol. 36, No. 20, 2003
Poly(ethylene oxide)-Lignin Blends 7809
Ta ble 6. IR Ba n d s (cm -1) of F r ee, In tr a m olecu la r , a n d In ter m olecu la r Hyd r oxyl Gr ou p s of Lign in Mod el Com p ou n d s
lignin model compounds
I
II
III
IV
V
phenol
3611(s)
free hydroxyl groupa
3639 (s)c
3616 (s)
intramolecular hydrogen bond
intermolecular hydrogen bondb (dimer)
intermolecular hydrogen bondb (multimer)
3556 (s)
3508 (s)
3444 (s)
3554 (s)
3490 (s)
3458 (s)
3552 (s)
nondetect
3219 (B)
3500 (B)
3390 (B)
3484 (B)
3416 (B)
nondetect
3226 (B)
a
b
IR band for free and intramolecular hydrogen bond was measured in diluted CCl4 solution (0.01 mol L-1). Intermolecular hydrogen
bonds were estimated from band center of nondiluted sample. The model compound was melt spread on a ZnSe plate at 150 °C for 5 min
under nitrogen atmosphere. c Key: (s) strong; (B) broad.
F igu r e 8. DR-FTIR spectra of lignin/PEO (A) and M-lignin/
PEO (B) blend fibers in the ν(C-O-C) asym region.
To resolve and further interpret the observed differences
in hydrogen bonding between the two systems (lignin/
PEO vs M-lignin/PEO), FT-IR studies were performed
F igu r e 7. DR-FTIR spectra of lignin/PEO (A) and M-lignin/
PEO (B) blends in the νOH region.
using various lignin model compounds. Table 6 lists the
νO-H bands for various lignin model compounds with
differing phenolic and aliphatic hydroxyl groups. Infra-
red bands for “free” or non-hydrogen-bonded hydroxyl
stretching modes were only observed in dilute solutions
of the aliphatic hydroxyl group containing model com-
pounds I and II; the phenolic compounds (III-V) form
a strong intramolecular hydrogen-bond with the adja-
cent methoxyl group even under dilute solution condi-
tions. At high concentration (Figure 9A), and in the solid
state, all of the lignin model compounds formed various
intermolecular hydrogen-bonding complexes (dimer
∼3500 cm-1 and multimer ∼3400 cm-1).
of the blends. As shown in Figure 8A, the νas(C-O-C) band
of lignin and PEO appear at the same position (1116
cm-1). However, the observed νas(C-O-C) band center
appears at a lower wavenumber in the lignin/PEO 50/
50 (1111 cm-1) and 25/75 (1109 cm-1) blends, consistent
with the observed effect in the νO-H region. The νas(C-O-C)
band of the M-lignin/PEO blends is shown in Figure 8B.
Unlike the original lignin, the νas(C-O-C) band of the
M-lignin appears at a higher wavenumber, 1126 cm-1
relative to that of PEO. Moreover, there appears to be
no significant shift in the band center associated with
this region with increasing blend ratio. Therefore, the
large change in the νO-H region of the lignin/PEO blends
relative to the M-lignin/PEO blends, along with the
differences between the νas(C-O-C) regions, suggests the
formation of strong hydrogen bonds between the phe-
nolic hydroxyl groups and PEO.
Blending the various lignin model compounds with
PEO revealed distinct differences in hydrogen-bonding
patterns (Figure 9B). Compounds I and II exhibited
similar band envelopes centered at ∼3450-3490 cm-1
.
These spectra showed slight differences to that observed
for the lignin model compounds in the absence of PEO
(Figure 9A); the band envelope sharpened, indicating a
less complex hydrogen-bonding system. The absence of
a shift in wavenumber indicates the strength of the
hydrogen bond between aliphatic hydroxyl groups such
The band envelopes of the lignin, M-lignin and
corresponding blends (Figure 7) are a complex super-
position of various types of OH- - -O hydrogen bonds.

7810 Kadla and Kubo
Macromolecules, Vol. 36, No. 20, 2003
as a result of dilution with the synthetic polymer FT-
IR analysis of lignin model compound/polystyrene (PS)
blends (Figure 9C) were performed. Unlike the PEO
blends (Figure 9B), incorporation of PS produces a very
broad band envelope for compounds I and II, whereas
for III, IV, and V very sharp νO-H bands are observed.
In fact, the FT-IR spectra of the various lignin model
compound blends with PS (Figure 9C) are identical to
FT-IR spectra obtained from similar concentration
lignin model compound solutions in carbon tetrachloride
(Figure 9A).
Purcell and Drago25 have shown that the wavenumber
shift (∆νO-H) between the free hydroxyl stretching
vibration and that of hydrogen-bonded species yields a
measure of the average strength of the intermolecular
interactions. From the lignin model compounds inves-
tigated it can be seen that very strong intermolecular
interactions exist between the various hydroxyl groups
in lignin (Table 6) as well as in lignin/PEO blends. In
fact, in the pure lignin and the M-lignin/PEO blends
∆νO-H ≈ 189 cm-1, whereas in the lignin/PEO blends
∆νO-H increases to about 316 cm-1 (a corresponding
wavenumber shift in the C-O-C stretching region in
PEO is also observed, ∆νC-O-C ≈ 10 cm-1). This suggests
that the average hydrogen bond between the ether
oxygen in PEO and the phenolic hydroxyl groups in
lignin is stronger than with the aliphatic hydroxyl
groups and that the hydrogen-bonds involving aliphatic
hydroxyl groups are similar to those within the lignin
homopolymer itself.
Su m m a r y
Lignin-based thermoplastic blends have been pro-
duced using a commercial technical lignin and PEO.
Incorporation of small amounts of PEO (5-10% w/w)
sufficiently disrupts the lignin supramacromolecular
complexes leading to enhanced physical properties.
Increasing PEO incorporation further disrupts the su-
pramacromolecular lignin structure and the observed
physical properties become more influenced by the PEO
component. Miscible blends were observed over the
entire blend ratio. The basis for the miscibility is
hydrogen-bonding interactions between the hydroxyl
protons of lignin and the proton accepting sites of PEO.
The strength of the hydrogen-bonding interactions
observed is different for the aromatic and aliphatic
hydroxyl groups in lignin. FT-IR analyses revealed a
strong hydrogen bond between the aromatic hydroxyl
proton and the ether oxygen in PEO; no such hydrogen
bond was observed in the lignin involving only aliphatic
hydroxyl group hydrogen-bonding. However, a negative
deviation of Tg from the weight-average values was
observed in both blend systems, suggesting a polymer
blend system involving weak specific intermolecular
interactions. This apparent contradiction may be ex-
plained by the following: (i) the number or extent of
hydrogen-bonding present may be relatively low, as
suggested by the Kwei fitting parameters, and/or (ii)
although strong intermolecular hydrogen-bonding ex-
ists, the effect of PEO incorporation on blend Tg may
depend on the nonbonding interactions that are respon-
sible for the supramacromolecular complexes present in
lignin-based systems.
F igu r e 9. FTIR spectra of lignin model compounds collected
in CCl4 (A), PEO (B), and PS (C) in the νOH region.
as those found in lignin (I and II) and PEO are of the
same magnitude as those between such groups within
lignin itself. These results are consistent with, and
support the findings wherein the band envelope of the
νO-H region for the M-lignin/PEO blends (Figure 7B) did
not substantially change with the incorporation of PEO.
The monoaromatic phenolic compounds III and IV
had similar band envelopes centered at ∼3420 cm-1
(Figure 9B). However, unlike compounds I and II, the
band envelopes of III and IV are much broader and
centered at a lower wavenumber in the presence of PEO.
This result suggests a change in hydrogen-bonding in
which a stronger, more complex hydrogen-bonding
system is formed. Interestingly, the addition of PEO to
compound V showed not only a shift in wavenumber,
but also a sharpening of the band envelope; two discern-
ible band centers at 3226 and 3146 cm-1, respectively,
are present.
FT-IR analysis of the lignin model compounds shows
the presence of strong hydrogen bonds between the
various hydroxyl groups present (Figure 9A). PEO
(Figure 9B) appears to interact with the phenolic
hydroxyl moieties (compounds III-V) to a greater and
different extent than that of the aliphatic hydroxyl
moieties (compounds I and II), consistent with the
lignin/PEO and M-lignin/PEO blends. To confirm the
observed change in the hydroxyl stretching bands upon
polymer addition is due to specific interactions between
lignin and PEO and not the result of enhanced inter-
molecular interactions between lignin macromolecules
Ack n ow led gm en t . Financial support from North
Carolina State University and the US Department of
Agriculture is gratefully acknowledged.

Macromolecules, Vol. 36, No. 20, 2003
Poly(ethylene oxide)-Lignin Blends 7811
(22) Kwei, T. K. J . Polym. Sci., Part C: Polym. Lett. 1984, 22,
307-313.
Refer en ces a n d Notes
(1) Imam, S. H.; Greene, R. V.; Zaidi, B. R. Biopolymers: utilizing
nature’s advanced materials; American Chemical Society:
Washington, DC, 1999.
(23) Kwei, T. K.; Pearce, E. M.; Pennacchia, J . R.; Charton, M.
Macromolecules 1987, 20, 1174-1176.
(24) Kuo, S. W.; Chang, F. C. Macromolecules 2001, 34, 4089-
(2) Martuscelli, E.; Palumbo, R.; Kryszewski, M.; Galeski, A.
Polymer blends: processing, morphology, and properties:
proceedings of the First J oint Italian-Polish Seminar on
Multicomponent Polymeric Systems, held in Capri, Italy,
October 16-21, 1979; Plenum Press: New York, 1980.
(3) Feldman, D.; Banu, D.; Lacasse, M.; Wang, J .; Luchian, C.
J . Macromol. Sci.sPure Appl. Chem. 1995, A32, 1613-1619.
(4) Tudorachi, N.; Cascaval, C. N.; Rusu, M. J . Polym. Eng. 2000,
20, 287-304.
(5) Yamane, H. Nihon Reoroji Gakkaishi 1999, 27, 213-218.
(6) Glasser, W. G.; Sarkanen, S. Lignin: properties and materi-
als, American Chemical Society: Washington, DC, 1989.
(7) Meister, J . J .; Chen, M.-j. U.S. Patent 5,424,382, 1995.
(8) Li, Y.; Mlynar, J .; Sarkanen, S. J . Polym. Sci., Part. B: Polym.
Phys. 1997, 35, 1899-1910.
4097.
(25) Purcell, K. F.; Drago, R. S. J . Am. Chem. Soc. 1967, 89, 2874-
2880.
(26) Chen, C.-L. In Springer series in wood science; Dence, C. W.,
Ed.; Springer-Verlag: Berlin, New York, 1992; pp 301-321.
(27) Lin, S. Y. In Methods in lignin chemistry; Dence, C. W., Ed.;
Springer-Verlag: Berlin, New York, 1992; pp 217-232.
(28) Lundquist, K. In Methods in lignin chemistry; Lin, S. Y.,
Dence, C. W., Eds.; Springer-Verlag: Berlin, New York, 1992;
pp 242-249.
(29) Kubo, S.; Ishikawa, N.; Uraki, Y.; Sano, Y. Mokuzai Gakkai-
shi 1997, 43, 655-662.
(30) Hoffman, J . D.; Weeks, J . J . J . Chem. Phys. 1962, 37, 1723-
1738.
(31) Li, Y.; Sarkanen, S. Macromolecules 2002, 35, 9707-9715.
(32) Guo, Q. P.; Huang, J . Y.; Li, B. Y.; Chen, T. L.; Zhang, H. F.;
Feng, Z. L. Polymer 1991, 32, 58-65.
(33) Ikeda, T.; Holtman, K.; Kadla, J . F.; Chang, H. M.; J ameel,
H. J . Agric. Food Chem. 2002, 50, 129-135.
(34) Wolf, M.; Wendorff, J . H. Polym Commun 1990, 31, 226-
228.
(9) Wang, J . S.; Manley, R. S.; Feldman, D. Prog. Polym. Sci.
1992, 17, 611-646.
(10) Flory, P. J . Principles of polymer chemistry; Cornell Univer-
sity Press: Ithaca, NY, 1953.
(11) Paul, D. R.; Bucknall, C. B. Polymer blends; Wiley: New York,
2000.
(12) Wang, J .; Cheung, M. K.; Mi, Y. L. Polymer 2002, 43, 1357-
(35) Lin, A. A.; Kwei, T. K.; Reiser, A. Macromolecules 1989, 22,
1364.
4112-4119.
(13) Oulevey, F.; Burnham, N. A.; Gremaud, G.; Kulik, A. J .;
Pollock, H. M.; Hammiche, A.; Reading, M.; Song, M.;
Hourston, D. J . Polymer 2000, 41, 3087-3092.
(14) Ilzhoefer, J . R.; Spontak, R. J . Langmuir 1995, 11, 3288-
3291.
(15) Schantz, S. Macromolecules 1997, 30, 1419-1425.
(16) Couchman, P. R. Macromolecules 1978, 11, 1156-1161.
(17) Couchman, P. R. Macromolecules 1987, 20, 1712-1717.
(18) Couchman, P. R. Macromolecules 1991, 24, 5772-5774.
(19) Dimarzio, E. A.; Gibbs, J . H. J . Polym. Sci. 1959, 40, 121-
131.
(36) Brisson, J .; Breault, B. Macromolecules 1991, 24, 495-504.
(37) Nishi, T.; Wang, T. T.; Kwei, T. K. Macromolecules 1975, 8,
227-234.
(38) Cheung, Y. W.; Stein, R. S. Macromolecules 1994, 27, 2512-
2519.
(39) J onza, J . M.; Porter, R. S. Macromolecules 1986, 19, 1946-
1951.
(40) J o, W. H.; Lee, S. C. Macromolecules 1990, 23, 2261-2265.
(41) Iriarte, M.; Iribarren, J . I.; Etxeberria, A.; Iruin, J . J . Polymer
1989, 30, 1160-1165.
(20) Gordon, M.; Taylor, J . S. J . Appl. Chem. 1952, 2, 493-500.
(21) Lu, X.; Weiss, R. A. Macromolecules 1992, 25, 3242-3246.
MA0348371