- Email: [email protected]
A universal route towards thermoplastic lignin composites with improved mechanical properties
Polymer 55 (2014) 995e1003
Contents lists available at ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
A universal route towards thermoplastic lignin composites with improved mechanical properties Shayna L. Hilburg b, Allison N. Elder a, Hoyong Chung a, Rachel L. Ferebee b, Michael R. Bockstaller b, Newell R. Washburn a, c, * a b c
Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, United States Department of Materials Science & Engineering, Carnegie Mellon University, 5000 Forbes Ave, Pittsburgh, PA 15213, United States Department of Biomedical Engineering, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, United States
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 October 2013 Received in revised form 20 December 2013 Accepted 29 December 2013 Available online 8 January 2014
Nanocomposites based on synthetic polymers grafted from kraft lignin with average particle size of 5 nm were synthesized using atom transfer radical polymerization (ATRP). Lignin macroinitiators were prepared, and polystyrene and poly(methyl methacrylate) were polymerized with target degree of polymerization of 450 resulting in materials having lignin mass fractions of 4.5%, 8.3%, and 22.1% for the poly(methyl methacrylate) samples and 3.2%, 7.1%, and 19.6% for the polystyrene samples. Tensile testing showed a decreased modulus but enhanced toughness of all nanocomposites compared to homopolymers, and the poly(methyl methacrylate)-grafted samples had nearly twice the ultimate elongation as the polystyrene grafts at high graft density. Both types of grafted nanocomposites had toughness values that were greater than 10-times that of the corresponding kraft-lignin/polymer blend system, indicating the potential of ATRP as the basis for the ‘one component’ composite approach towards more sustainable polymeric materials. Dynamical mechanical analysis was used to measure softening temperatures, and both the polystyrene-grafted and poly(methyl methacrylate)-grafted nanocomposites had a peak in the loss modulus that was higher than the corresponding homopolymer, consistent with strong polymer elignin interactions. Lignin grafted with thermoplastic polymers could be an important material based on an inexpensive, renewable feedstock that offers unique mechanical properties compared with many other nanocomposites based on inorganic nanoparticles. Our results indicate that ATRP is well suited for preparing lignin-based thermoplastics and could be the basis for hybrid materials that make effective use of this important renewable resource. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Lignin Nanocomposite ATRP
1. Introduction The integration of natural-sourced polymers with synthetic commodity and engineering plastics is widely recognized to be an important cornerstone in the broader context of developing more sustainable and low-energy polymeric materials [1e8]. Kraft lignin e a term used to describe a range of polydisperse and branched phenolic polymers that are derived from native lignin via the kraft pulping process e has received particular attention due to the relevance of lignin as medium for solar energy storage in plants and its abundance as an important structural component of wood. The high thermal, oxidative, and hydrolytic [9] stability as well as favorable mechanical properties (the Young’s modulus has been
* Corresponding author. Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, United States. E-mail address: [email protected] (N.R. Washburn). 0032-3861/$ e see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.polymer.2013.12.070
estimated to 2 GPa [10]) have motivated interest in using kraft lignin as a filler, for example, for the reinforcement of elastomeric materials [11e18]. However, while the application of lignin in elastomeric as well as thermoset material systems has been widely reported [19], its application as filler in thermoplastic polymeric systems has been a challenge. Among the relevant nonpolar engineering polymer systems, best results were achieved by blending kraft lignin with poly(methyl methacrylate) (PMMA). For example, blends containing 5e20% hydroxypropyl lignin in PMMA were found to have single glass transition temperatures that were around 5 C below that of pure PMMA and a Young’s modulus of 1800e1900 MPa, depending on lignin molecular weight and mass fraction in the blend [20]. This value of Young’s modulus represented a 200 MPa increase over the pure polymer, however, all blends were found to be brittle with a strain-to-fracture less than 2%. The limited effect of kraft lignin was rationalized as a consequence of the poor interfacial binding between filler particles and the matrix that favors particle aggregation and phase separation in
996
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
lignin/polymer blends [21e26]. The development of novel synthetic strategies to facilitate miscible polymer/lignin blends is therefore an important challenge to realize the potential benefits of lignin filler materials for the development of thermoplastic polymer blends. The challenge of facilitating miscible kraft lignin dispersions bears similarity to the broader challenge of understanding the governing parameters that control miscibility in particle/polymer blends [23]. The latter systems have received particular attention in the context of inorganic particle/polymer dispersions. A widely used strategy to compatibilize particle fillers within polymeric hosts is by grafting of polymer chains that are chemically identical to the matrix polymer to the particle matrix [27,28]. In these athermal polymer/particle blends mixing is determined by the subtle interplay of entropic contributions that sensitively depend on the architecture of the polymer-grafted filler (i.e. the density and degree of polymerization of grafted chains as well as the particle size). In general, miscibility is expected only if the degree of polymerization of grafted chains is at least of the same order as the matrix [29]. As recently demonstrated by Ojha et al., [30] the requirement of high molecular grafted chains severely limits the attainable filler concentration and thus the blending of polymergrafted (lignin) fillers with the corresponding matrix polymers cannot be seen as a viable strategy towards lignin/polymer composites with high lignin content and improved property characteristics. An alternative strategy to attain high filler content polymer nanocomposites is based on the concept of ‘one-component’ hybrid materials e nanocomposite materials that are synthesized by the assembly of polymer-grafted particles [31]. Primary advantages of the ‘one-component’ approach are the fine control of the materials microstructure (note that miscibility gaps are absent in onecomponent systems) along with mechanical properties that resemble those of higher molecular weight polymers. Due to their technological relevance as commodity and engineering plastics as well as the accessibility to viable controlled radical polymerization techniques, polystyrene (PS) and PMMA have been studied extensively as model systems in one-component nanocomposites. For example, Choi et al. recently demonstrated that the assembly of PMMA- or polystyrene- (PS) grafted silica particles with suitable architecture enables the fabrication of nanocomposite materials with highly regular microstructure [32]. Interestingly, onecomponent nanocomposites were found to exhibit significantly increased fracture toughness (as compared to the corresponding binary particle/polymer blend systems) e this characteristic was explained as a consequence of the more effective entanglement structure in branched polymeric systems [33]. We note that a particular advantage of the ‘one-component’ approach towards nanocomposite materials is that it is applicable to any type of ‘matrix’ polymer as long as synthetic methodologies exist to facilitate the controlled grafting of the polymer to the particle system. The general concept of one-component nanocomposite materials is illustrated in Scheme 1. The objective of this article is to illustrate the application of ATRP in preparing thermoplastic ‘one-component composites’ that combine a high filling fraction of lignin with improved fracture resistance as compared to the binary polymer/lignin blend systems. The structure of this paper is as follows: In the first part we discuss the conditions that allow for application of atom-transfer radical polymerization to synthesize a set of PMMA- and PS-grafted lignin particle systems with systematic variation of the number and degree of polymerization of surface grafted chains. In the second part we discuss the effect of polymer graft modification on the mechanical properties of lignin-based one-component composites and we demonstrate that even in case of lowest density of polymer
Scheme 1. Comparison of ‘binary’ and ‘one-component’ nanocomposite fabrication schemes.
grafts per lignin ‘particle’ a significant improvement of fracture resistance as compared to ‘classical’ binary composites was observed. Thus, polymer-grafted lignin particles with appropriate architecture could present a novel platform for the fabrication of low-energy thermoplastic commodity polymers with high lignin content and enhanced mechanical properties.
2. Experimental section 2.1. Materials Lignin was purchased from Tokyo Chemical Industry, Co. Methyl methacrylate (MMA), styrene, ethyl 2-bromoisobutyrate (EBiB), CuCl, CuBr, N,N,N0 ,N0 ,N00 -pentamethyldiethylenetriamine (PMDETA), 2,20 -bipyridine (bpy), and pentafluorobenzaldehyde were purchased from SigmaeAldrich. Styrene and MMA monomers were run through a basic alumina column before use to remove inhibitor. All other chemicals were used as received.
2.2. ATRP of styrene First, styrene, dimethylformamide (DMF), PMDETA, and EBiB were dried over molecular sieves and degassed under nitrogen for 30 min. Then CuBr (14.8 mg, 0.103 mmol) was added to a 50 mL Schlenk flask, which was subsequently degassed by three vacuume nitrogen cycles. Next, 4.5 mL DMF, 21.5 mL (0.103 mmol) PMDETA, 5.31 mL (46.3 mmol) styrene, and 15.1 mL (0.103 mmol) EBiB were added to the flask and the reaction was carried out at 100 C for 7 days. The percent conversion was determined to be 52% using 1H NMR by comparing the disappearance of acrylate peaks to DMF internal standard. Copper was removed by dilution with THF and filtration through neutral alumina. The polymer was precipitated into 100 mL of hexane and dried in vacuo at room temperature. The resulting polymer (1.2401 g, 51.4% yield) was a white brittle solid.
2.3. ATRP of methyl methacrylate ATRP of MMA was conducted in the same manner as described above. First, MMA, DMF, PMDETA, and EBiB were bubbled under nitrogen for 30 min and 17 mg (0.119 mmol) of CuBr was added to a Schlenk flask with three vacuumenitrogen cycles. MMA (5.03 mL, 47.2 mmol), 1.2 mL DMF, PMDETA (21.9 mL, 0.105 mmol), and EBiB (15.4 mL, 0.105 mmol) were added and the reaction was carried out at 70 C for 24 h. Percent conversion according 1H NMR was 62%. Again, copper was removed by dilution with THF and filtration through neutral alumina. A total of 1.9543 g (66.7% yield) was obtained after precipitation into hexane and drying in vacuo.
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
997 OH
OH
Lignin
O HO
Lignin
O Br
Br
HO
O Pyridine, THF
H or Lignin
H3CO
R.T. 24 hours
H or Lignin
H3CO
O
OH
O
Br
Scheme 2. Synthesis of lignin macroinitiator.
2.4. Lignin acidification
2.6. ATRP of styrene-grafted lignin samples
Lignin (5.00 g) was suspended in 75 mL of 2 M HCl with stirring for 15 min. The mixture was vacuum filtered and the acidification/ filtration process was repeated. During the second filtration, the solid lignin was rinsed with water and then allowed to air-dry overnight. Finally, the solid powder was dried under vacuum for 5 h at 35 C. The total amount obtained was 3.30 g giving a 66% yield.
For polymer graft synthesis using the macroinitiator with highest concentration of initiators, 0.3082 g lignin macroinitiator, 21.0 mg CuCl (0.212 mmol), and 58.2 mg bpy (0.373 mmol) were added to the Schlenk flask and degassed using three vacuumenitrogen cycles. DMF and styrene were bubbled under nitrogen for 30 min then DMF (8.5 mL) and styrene (10.4 mL, 90.8 mmol) were added to the flask. The reaction was carried out at 100 C for up to 14 days, which provided the highest reproducibility. However, most reactions can be terminated after a few days. The percent conversion from 1H NMR was determined to be 97.2%. The polymer was then diluted with THF, precipitated into hexane, and dried under vacuum at room temperature. Additional styrene-grafted polymers were synthesized using the same procedure. These polymers consisted of styrene grafted from lignin with lignin fractions of 4.5%, 8.3%, and 22.1%, and the amount of monomer, ligand, and catalyst were scaled in accordance with the moles of initiator. Scheme 3 presents a summary of the polymerization conditions.
2.5. Lignin macroinitiator synthesis Acidified lignin (1.06 g) was dissolved in 20 mL of pyridine in a 200 mL roundbottom flask. The lignin solution was bubbled under nitrogen for 30 min, followed by the addition of 10 mL of deoxygenated THF. To prepare samples with the highest polymer content (4.5% lignin), 186 mL (0.261 g, 1.14 mmol) of a-bromoisobutyryl bromide was added dropwise. The flask was then allowed to stir for 24 h at room temperature. After 24 h, the lignin was precipitated into 400 mL diethyl ether. The ether was decanted off and the resulting solid was dried at 35 C under vacuum. The total amount of initiator obtained was 1.17 g (88% yield). The number of initiator sites per mass of lignin was determined by 1H NMR in DMSO with 10 mL of pentafluorobenzaldehyde (PFB) as an internal standard. 1H NMR (300 MHz) d ¼ 2.0 ppm (s, 6 H, conjugated initiator), d ¼ 10.19 ppm (s, 1 H, PFB). Additional lignin initiators with lower concentrations of initiator were synthesized using the same procedure, but scaling down the a-bromoisobutyryl bromide accordingly. A summary of the conditions for lignin macroinitiator synthesis is shown in Scheme 2.
2.7. ATRP of methyl methacrylate-grafted lignin samples For polymer graft synthesis using the macroinitiator with the highest polymer:lignin ratio, 0.1915 g initiator (ca 0.126 mmol), 18.8 mg CuBr (0.131 mmol), and 26.3 mL PMDETA (0.126 mmol) were added to the Schlenk flask and degassed using three vacuume nitrogen cycles. DMF and MMA were bubbled under nitrogen for 30 min, then DMF (7.5 mL) and MMA (6.0 mL, 56.3 mmol) were added to the flask. The reaction was carried out at 70 C for 24 h.
OH
OH
Lignin
O
Lignin
O
HO
HO
H or Lignin
H3CO O
H or Lignin
CuCl, bpy, DMF 100°C
O
O
O
Br
n
Scheme 3. Synthesis of styrene-grafted lignin.
998
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003 OH
OH
Lignin
O
Lignin
O
O
HO
HO
O
H or Lignin
H3CO
H or Lignin
PMDETA, CuBr, DMF 70°C
O
O O
O
Br
O n
O
Scheme 4. Reaction conditions for synthesis of MMA-grafted lignin.
The percent conversion from 1H NMR was determined to be 72.3%. The polymer was then diluted with THF, precipitated into hexane, and dried under vacuum at room temperature. Additional MMA-grafted polymers with lower polymer:lignin ratios were synthesized using the same procedure. To ensure high conversion, reactions were run up to 8 days, but tuning reaction conditions can increase rates. Scheme 4 presents a summary of the reaction conditions for MMA-grafted lignin. 2.8. NMR NMR measurements were carried out using a Bruker Advance 300 MHz NMR Instrument in CDCl3 or deuterated dimethyl sulfoxide. 2.9. Tensile testing
(a)
Tensile stress-strain tests were performed on an Instron tensile test machine using pneumatic grips and operated at a crosshead speed of 0.15 mm/min. Specimen for tensile test evaluation were prepared by first hotpressing of samples at 120 C and subsequent cutting of strips with test region dimensions of 20 mm 5 mm 1.5 mm. 2.10. Dynamic mechanical analysis (DMA) Disks of approximately 8 mm in diameter and 2 mm height were prepared from each sample. Polymer was pressed into a cylindrical mold at 120 C then annealed at 100 C under vacuum for at least 48 h. DMA was performed on a TA Instruments RSA-G2 using 15 mm disk compression clamps. Samples were equilibrated at 150 C before oscillating at 0.25% strain at 1 Hz while decreasing temperature at 1 C/min.
(b)
3. Results and discussion 3.1. Chemical and structural analysis Representative NMR data for unmodified lignin, lignin modified with the ATRP initiator, and MMA-grafted lignin (4.5% lignin) are shown in Fig. 1(a)e(c), respectively. Due to lignin being present in a very small mass fraction (<5%), the lignin peaks for the grafted material are difficult to see. Pentafluorobenzaldehyde was added to the sample shown in Fig. 1(b) to allow for quantitative measurement of the moles of initiator per gram of lignin. Additional peaks in the grafted material shown in Fig. 1(c) appear at 7.9, 2.7, and 2.8 ppm and are due to residual DMF from the polymerization
(c) Fig. 1. 1H NMR of lignin (a), lignin macroinitator (b), and lignin-g-PMMA (c).
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
999
Table 1 Compositions of samples reported in this study. Sample
Polymer graft
Average DP
Mass % lignin
lignin22.1PMMA401 lignin8.3PMMA388 lignin4.5PMMA323 lignin19.6PS449 lignin7.1PS444 lignin3.2PS437
PMMA PMMA PMMA PS PS PS
401 388 323 449 444 437
22.1 8.3 4.5 19.6 7.1 3.2
reaction and at 1.4 due to hexane. Based on quantification of the spectra, the initiator site concentrations for the three macroinitiators were measured to be 80, 280, and 640 mmol/(g lignin). Based on a GPC-derived lignin molecular weight of 25,000 g/mol, this corresponds to grafting densities of approximately 2, 7, and 16 per lignin molecule. However, initiator site density will be reported based on mass of lignin instead of per lignin particle because of the high level of heterogeneity in lignin. While it is a three-dimensional polymer, in using it as a macroinitiator we treat it similarly to inorganic nanoparticles, which are represented on a “per mass” basis instead of “per molecule” in one-component nanocomposites. The solubility of the products indicates polymerization from the lignin surface has been performed. Attachment of initiators offers high monomer conversion and essentially quantitative grafting efficiency compared with other methods, such as radiationinduced polymerization or reacting lignin with methacryloyl chloride, [34] in providing a means for continuously varying composition without forming crosslinked products. Reactive blending of lignin generally results in less than 100% grafting efficiency even using isocyanate-terminated polymers, suggesting that grafting-from chemistries provide a more efficient means for surface modification [35]. Indeed, Kim and Kadla demonstrated the utility of ATRP in polymerization of N-isopropylacrylamide from lignin macroinitiators to prepare thermal responsive materials [36]. A similar approach was taken by Tang and co-workers in the preparation of hydrophobic polymer-grafted lignin by ATRP of rosin polymers from a lignin macroinitiator [37]. The current research extends these approaches to making high-strength one-component composites, demonstrating ATRP offers significant advantages over other approaches taken to synthesize lignin grafted with thermoplastic polymers. The compositions of the six polymer-grafted lignin nanocomposites synthesized in this work are summarized in Table 1. Samples are identified as ligninxxPOLYMERyyy where xx is the mass fraction of lignin in the sample, POLYMER is PMMA or PS, and yyy is the average degree of graft polymerization based on monomer conversion. All samples were dark brown in color and dissolved readily in pyridine or DMF. 3.2. Structural characterization In Fig. 2 are shown the (intensity-averaged) distribution of hydrodynamic radii (RH) for the PMMA homopolymer, kraft lignin, and lignin4.5PMMA323 that were obtained by DLS analysis in DMF solution. The peak in the size distribution of lignin was 5 nm, PMMA homopolymer was 7 nm, while the value for the PMMAgrafted lignin was 28 nm. There is an aggregation in the dissolved copolymers, evidenced by a peak at larger particle size, but not to a large extent. Note also that the shape of the distribution of hydrodynamic radii is consistent with the expected heterogeneity of the polymer-conjugated product [38] e in particular the absence of a shoulder at small RH suggests that ATRP initiated cleanly from the lignin macroinitiator.
Fig. 2. Dynamic light scattering data from DMF solutions of kraft lignin (top), PMMA homopolymer (middle), and lignin4.5PMMA323 (bottom).
The conclusion of successful graft modification is also confirmed by transmission electron microscopy of solution-cast films of polymer-grafted lignin particles. Films were deposited from a 1 mg/ mL pyridine solution onto TEM grids and stained with OsO4 (selective for lignin) to provide contrast between lignin and PMMA. Fig. 3a depicts a representative TEM image of a thin film of lignin4.5PMMA323 revealing a uniform dispersion of lignin particles (dark) in PMMA matrix. Interestingly, the average nearest-neighbor particle distance hdi w 15 nm that can be determined from the micrograph (see inset of Fig. 3a) is consistent with the presence of PMMA grafts with degree of polymerization N ¼ 323 for which the radius of gyration of surface-grafted chains can be estimated to Rg,PMMA323 w 7 nm. To assess the implication of polymer graft modification on the mechanical properties of lignin composites, reference systems comprised of binary blends of lignin dispersed within the respective host matrix were prepared. For example, Fig. 3b depicts a representative electron micrograph of a binary lignin/PMMA blend with equal composition (i.e. lignin content) as the one-component lignin composite shown in Fig. 3a. The figure reveals a dispersed particle morphology in which the average diameter of particles is found to be of order 10 nm, similar to what was measured using DLS for individual lignin particles. Images of PMMA-grafted materials with lower graft densities appeared to have greater direct ligninelignin interactions, consistent with the mass fraction of lignin increasing from 4.5% to 8.3% and 22.1%, although the higher staining levels made it difficult to clearly observe the morphology (images not shown). The miscibility of lignin in PMMA may be understood as a consequence of the compatibilization effect that is imparted by hydroxyl functionalities on the surface of lignin that allow for hydrogen-bond formation with the PMMA host. Note that both systems e the one-component lignin4.5PMMA323 as well as the binary lignin/PMMA blend e are equal in morphology and composition and therefore are ideally suited to identify the relevance of molecular connectivity on the properties of the resulting composite material. We also note that lignin was found to be incompatible in PS host matrices, for which aggregated morphologies were observed (results not shown here). 3.3. Mechanical and thermal characterization The mechanical properties of the PMMA-lignin nanocomposites were characterized using tensile testing at room temperature.
1000
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
the effect of lignin content. Since PMMA and PS nanocomposites with lowest grafting densities had similar stressestrain curves, with ultimate elongations of 20% and tensile strengths of 5 MPa, we conclude that at higher lignin content the mechanical properties of the one-component nanocomposites were primarily determined by the lignin component. Interestingly, one-component lignin-g-PMMA systems were found to significantly exceed the fracture toughness of the corresponding binary blend systems. Lignin4.5PMMA323 had a toughness of 9.6 MJ/m3, which was more than 10-times greater than that of the PMMA homopolymer or the binary lignin-PMMA blend reference system. The improved toughness and ductility of onecomponent systems is consistent with recent reports on the mechanical properties of one-component composites based on PMMA-grafted silica particle systems that also suggest an enhancement of fracture toughness in one-component composite systems as compared to the reference binary blend systems [33]. In analogy to the inorganic homologues we interpret the increase of fracture toughness to be a consequence of the more efficient connectivity among entanglement points in the star-polymer-like lignin4.5PMMA323. The latter should support the interaction between entanglement points during craze formation and hence raise the energy dissipation during fracture. The non-continuous trend of modulus and toughness that is observed in both polymer-grafted lignin one-component composite materials with increasing number of polymer is surprising and counterintuitive, however, since in both cases the best performance was observed for
Fig. 3. Representative bright-field TEM images of lignin4.5PMMA323 composite materials. Panel a: lignin4.5PMMA323 one-composite revealing uniform particle dispersion. Scale bar is 200 nm. Inset shows distribution of next-nearest neighbor distances. Average distance is determined as hdi z 25 nm consistent with the presence of PMMA grafts with N ¼ 323. Panel b: binary lignin/PMMA blend system with equal volume fraction lignin as system shown in (a). Micrograph reveals uniform lignin-particle dispersion in PMMA. The scale bar is 500 nm. Inset shows magnified view of highlighted area depicting individual lignin particles dispersed within PMMA matrix. Scale bar is 20 nm.
Stressestrain curves were recorded for each polymer-grafted sample and compared with homopolymers of PMMA and PS having similar average molecular weight as the polymer grafts as well as the 1:7 binary blend reference system of lignin and PMMA homopolymer. Representative curves are shown in Fig. 4(a) and (b) for the grafted nanocomposites. The PMMA samples had ultimate elongation and toughness values that were equal to or greater than the values for the PS samples. This is consistent with previous results that have related the toughness of PMMA to the presence of a b-relaxation process associated with the motion of side chains [39]. Similarly, lignin4.5PMMA323 was found to exhibit a toughness value about twice the value of lignin3.2PS437. This likely reflects both intrinsic differences in the mechanical properties of PMMA and PS as well as
Fig. 4. Stress versus strain curves for (a) lignin-g-PMMA and (b) lignin-g-PS.
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
the lowest and highest concentrations of polymer grafts, respectively. We note that similar trends are observed for both material systems and thus we consider the result to be reliable. We hypothesize that the varying effect of polymer graft modification is a consequence of the competing influence of distinct governing parameters (such as the lignin volume fraction, the degree of polymerization and the conformation of grafted chains) on the mechanical properties of polymer-grafted lignin composite materials. Note that homopolymers and ligninehomopolymer blends were found to exhibit similar mechanical properties for PS and PMMA. The PS homopolymer had a modulus of 2700 MPa while the PMMA homopolymer was only 280 MPa, but the toughness values were significantly lower than that of high molecular weight analogues [40,41]. Furthermore, while the ultimate elongation of the PS was 5%, similar to what would be expected from a brittle polymer, that of PMMA was 15%, suggesting the latter is less brittle at lower molecular weights [39]. Solvent blending of lignin with each homopolymer resulted in modest increases in both modulus and toughness, although the magnitude of the increases was lower than what previously reported values [24,26,42,43]. Other researchers have demonstrated significant increases in modulus in particular, and the lack of increase may be due to the relatively low molecular weight of homopolymers used in the present study. Dynamical mechanical analysis (DMA) was performed on the samples to determine the softening temperatures of the samples, which can be associated with a glass transition. However, preliminary differential scanning calorimetry measurements on the grafted materials only showed broad thermal transitions (data not shown), suggesting these materials may not be true glass formers. Further study is necessary to clarify the physical nature of the softening transitions observed in DMA, but the peak in the loss modulus is reported as Tg for these purposes. Storage and loss moduli were taken from each of the six polymer samples, and representative plots for each sample are shown in Fig. 5. Samples were equilibrated at 150 C for 10 min, and data were collected as the samples cooled were cooled at 1 C/min. The storage modulus rises sharply as the samples cool through Tg, and the loss modulus exhibits the expected peak associated with this thermal transition. The trend in storage modulus for both PMMAand PS-grafted materials reflects that observed in Young’s modulus. Interestingly, Fig. 5 reveals that in the limit of small number of grafted chains the relaxation appears broadened both in case of PMMA and PS grafted lignin. This is consistent with the observed increase in toughness in case of a small number of polymer grafts and indicates a complex relationship between the architecture of grafted lignin and the mechanical properties of one-component lignin composites. The pure PMMA and PS samples had Tg values of 65 C and 80 C, respectively, while the 1:7 lignin blends had softening temperatures of 92 C and 95 C, respectively. These curves are omitted in Fig. 5 for clarity but had similar shapes as the grafted materials. The significant increase in Tg for the lignin-PMMA blend compared to the PMMA homopolymer suggests strong interactions between the nanoparticle and polymer matrix. The increase in Tg across all compositions is consistent with other measurements of polymer-grafted nanoparticles and is generally expected due to the constraints imparted on chain relaxation by bonding of one chain end to the inorganic particle. For example, silica-PS nanocomposites had Tg values that averaged 13 C greater than the corresponding homopolymer, which was attributed to chain confinement on the nanoparticle surface [44]. However, the 27 C increase is greater than that observed by Ciemniecki and Glasser, [20] who observed little or no difference in Tg when blending hydroxypropyl lignin into PMMA using tetrahydrofuran compatibilization, but this too may be due to chemically
1001
Fig. 5. Storage (black) and loss (gray) moduli for (a) lignin-g-PMMA samples and (b) lignin-g-PS samples with varying numbers of initiator sites per lignin.
functionalized lignin used in their study as well as differences in solvent processing. Table 2 summarizes the material parameters that were obtained through data fitting of tensile testing and DMA characterization. The toughening observed in these grafted one-component nanocomposites contrasts with the behavior observed in many ligninepolymer blends. Blends of up to 40% hydroxypropyl lignin in PMMA, investigated by Ciemniecki and Glasser, resulted in materials with a phase-separated morphology but one that displayed a single Tg that was insensitive to the lignin content [20]. The tensile strength of these blends varied from 50 MPa to 87 MPa depending on the lignin formulation and blending conditions, but the ultimate elongation values decreased relative to the unfilled polymer upon blending with even 5% lignin. In a separate study by Pucciariello et al., blends containing 10% straw lignin in PS having Mw of 280,000 showed a tensile strength of 12.0 MPa and an ultimate elongation of 11.8%, although the latter parameter decreases rapidly upon increasing the lignin content, suggesting this approach is not capable of increasing the amount of a renewable feedstock in PS [45]. One formulation with significantly different mechanical
1002
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
Table 2 Young’s modulus, toughness, ultimate elongation, and glass transition temperature values for PMMA and PS materials. Polymer
Young’s modulus (MPa)
lignin22.1PMMA401 lignin8.3PMMA388 lignin4.5PMMA323 PMMA (25 kDa) lignin/PMMA blend (1:7) lignin19.6PS449 lignin7.1PS444 lignin3.2PS437 PS (25 kDa) lignin/PS blend (1:7)
56 46 55 100 121 62 34 81 3100 2700
1 26 3 14 23 13 11 25 132 210
Toughness (MJ/m3) 1.0 2.7 3.7 0.15 0.24 0.37 0.28 0.52 0.012 0.014
0.3 1.4 0.1 0.1 0.1 0.25 0.05 0.24 0.002 0.002
Ultimate elongation (%) 23 58 52 12 13 12 15 14 3 4
4 26 4 5 3 7 1 7 1 1
Tg ( C)
77 75 82 65 92 79 103 96 80 95
properties was a blend containing 85% kraft lignin in a poly(vinyl acetate) matrix [25]. By carefully controlling the molecular weight of the lignin through a pH-dependent precipitation process, Sarkanen and co-workers were able to produce a blend having high lignin content that displayed an ultimate elongation of greater than 60% and a tensile strength of nearly 10 MPa [25]. However, the thermodynamic compatibility of kraft lignin and poly(vinyl acetate) was an important design principle, and poly(vinyl acetate) is not commonly used as a structural material in part due to its susceptibility to hydrolysis and conversion to poly(vinyl alcohol). The pronounced increase in fracture toughness that is realized in polymer-grafted lignin composites, while largely retaining the materials stiffness, renders the ‘one-component’ approach an appealing strategy in the fabrication of lignin-based thermoplastic engineering polymers.
4. Conclusion ATRP is well suited for preparing lignin-based nanocomposites with grafted thermoplastic polymers. Characterization using DLS confirmed that polymerization proceeds cleanly from the lignin macroinitiators and results in products defined by a lignin core with a polymer corona. Electron microscopy images of the ligning-PMMA sample having lignin mass fraction of 4.5% showed that the microstructure was based on a lignin core in a continuous matrix of PMMA. Tensile testing and DMA were used to characterize the mechanical properties and thermodynamic properties of lignin-g-PS and lignin-g-PMMA nanocomposites. The increase of Tg as compared to the homopolymer systems is consistent with the restriction of grafted polymer chains and indicates that interactions between lignin and both polymers is strong, but particularly in the lignin-g-PMMA samples, where the highest toughness values were observed. However, the highest moduli were observed for the compositions with the highest lignin content, suggesting the properties become less dependent on the specific polymer and more dependent on the intrinsic lignin properties. One-component composites based on polymer-grafted lignin were found to exhibit significantly enhanced fracture toughness as compared to binary lignin/polymer blends while retaining the stiffness characteristics of polymer glasses. The fortuitous combination of mechanical properties along with the fine control of the composite microstructure should render polymer-grafted lignin composites a platform for developing advanced materials based on renewable resources. This is particularly true since the current progress in controlled radical polymerization suggest the application of the approach to a wide range of lignin-based nanocomposites.
Acknowledgments Dr. Ludwig Leibler is gratefully acknowledged for helpful discussions. NMR instrumentation at CMU was partially supported by NSF (CHE-0130903 and CHE-1039870). R.L.F. and M.R.B. acknowledge support through the NSF-IGERT program via NSF-0966227. A.N.E. gratefully acknowledges the Richard King Mellon Presidential Fellows program for support.
References [1] Bataille P, Ricard L, Sapieha S. Effects of cellulose fibers in polypropylene composites. Polym Compos 1989;10:103e8. [2] Bledzki AK, Gassan J. Composites reinforced with cellulose based fibres. Prog Polym Sci 1999;24:221e74. [3] Chung H, Washburn NR. Chemistry of lignin-based materials. Green Mater 2012;1:137e60. [4] Feldman D. Lignin polymer systems and some applications. Prog Polym Sci 1986;12:271. [5] Li Y, Sarkanen S. Alkylated kraft lignin-based thermoplastic blends with aliphatic polyesters. Macromolecules 2002;35:9707e15. [6] Lundquist L. Life cycle engineering of plastics: technology, economy, and the environment. Oxford; New York: Elsevier; 2000. [7] Ratto JA, Stenhouse PJ, Auerbach M, Mitchell J, Farrell R. Processing, performance and biodegradability of a thermoplastic aliphatic polyester/starch system. Polymer 1999;40:6777e88. [8] Rodriguez-Gonzalez FJ, Ramsay BA, Favis BD. High performance LDPE/thermoplastic starch blends: a sustainable alternative to pure polyethylene. Polymer 2003;44:1517e26. [9] Sun Y, Cheng JY. Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresour Technol 2002;83:1e11. [10] Cousins WJ. Elastic-modulus of lignin as related to moisture-content. Wood Sci Technol 1976;10:9e17. [11] Cateto CA, Barreiro MF, Rodrigues AE, Brochier-Solan MC, Thielemans W, Belgacem MN. Lignins as macromonomers for polyurethane synthesis: a comparative study on hydroxyl group determination. J Appl Polym Sci 2008;109:3008e17. [12] Chung HY, Washburn NR. Improved lignin polyurethane properties with Lewis acid treatment. Acs Appl Mater Inter 2012;4:2840e6. [13] Hirose S, Hatakeyama T, Hatakeyama H. Synthesis and thermal properties of epoxy resins from ester-carboxylic acid derivative of alcoholysis lignin. Macromol Symp 2003;197:157e69. [14] Morck R, Reimann A, Kringstad KP. Elastomeric polyurethanes from a kraft lignin polyethylene-glycol toluene diisocyanate system. Acs Sym Ser 1989;397:390e401. [15] Nakano J, Izuta Y, Orita T, Hatakeyama H, Kobashigawa K, Teruya K, et al. Thermal and mechanical properties of polyurethanes derived from fractionated kraft lignin. Sen-I Gakkaishi 1997;53:416e22. [16] Reimann A, Morck R, Yoshida H, Hatakeyama H, Kringstad KP. Kraft lignin in polyurethanes. 3. Effects of the molecular-weight of Peg on the properties of polyurethanes from a kraft lignin Peg Mdi system. J Appl Polym Sci 1990;41: 39e50. [17] Thring RW, Vanderlaan MN, Griffin SL. Polyurethanes from Alcell (R) lignin. Biomass Bioenergy 1997;13:125e32. [18] Yoshida H, Morck R, Kringstad KP, Hatakeyama H. Kraft lignin in polyurethanes. 1. Mechanical-properties of polyurethanes from a kraft lignin polyether triol polymeric Mdi system. J Appl Polym Sci 1987;34:1187e98. [19] Lora JH, Glasser WG. Recent industrial applications of lignin: a sustainable alternative to nonrenewable materials. J Polym Environ 2002;10:39e48. [20] Ciemniecki SL, Glasser WG. Multiphase materials with lignin. 1. Blends of hydroxypropyl lignin with Poly(Methyl methacrylate). Polymer 1988;29: 1021e9. [21] Chen F, Dai HH, Dong XL, Yang JT, Zhong MQ. Physical properties of ligninbased polypropylene blends. Polym Compos 2011;32:1019e25. [22] Feldman D, Banu D, Manley RSJ, Zhu H. Highly filled blends of a vinylic copolymer with plasticized lignin: thermal and mechanical properties. J Appl Polym Sci 2003;89:2000e10. [23] Kadla JF, Kubo S. Lignin-based polymer blends: analysis of intermolecular interactions in lignin-synthetic polymer blends. Compos Part A-Appl S 2004;35:395e400. [24] Kubo S, Kadla JF. The formation of strong intermolecular interactions in immiscible blends of poly(vinyl alcohol) (PVA) and lignin. Biomacromolecules 2003;4:561e7. [25] Li Y, Mlynar J, Sarkanen S. The first 85% kraft lignin-based thermoplastics. J Polym Sci Pol Phys 1997;35:1899e910. [26] Wang JS, Manley RS, Feldman D. Synthetic-polymer lignin copolymers and blends. Prog Polym Sci 1992;17:611e46. [27] Rong MZ, Zhang MQ, Zheng YX, Zeng HM, Walter R, Friedrich K. Structureproperty relationships of irradiation grafted nano-inorganic particle filled polypropylene composites. Polymer 2001;42:167e83.
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003 [28] Balazs AC, Emrick T, Russell TP. Nanoparticle polymer composites: where two small worlds meet. Science 2006;314:1107e10. [29] Green PF. The structure of chain end-grafted nanoparticle/homopolymer nanocomposites. Soft Matter 2011;7:7914e26. [30] Ojha S, Dang A, Hui CM, Mahoney C, Matyjaszewski K, Bockstaller MR. Strategies for the synthesis of thermoplastic polymer nanocomposite materials with high inorganic filling fraction. Langmuir: ACS J Surfaces Colloids 2013;29:8989e96. [31] von Werne T, Patten TE. Preparation of structurally well-defined polymernanoparticle hybrids with controlled/living radical polymerizations. J Am Chem Soc 1999;121:7409e10. [32] Choi J, Dong H, Matyjaszewski K, Bockstaller MR. Flexible particle array structures by controlling polymer graft architecture. J Am Chem Soc 2010;132:12537e9. [33] Choi J, Hui CM, Pietrasik J, Dong HC, Matyjaszewski K, Bockstaller MR. Toughening fragile matter: mechanical properties of particle solids assembled from polymer-grafted hybrid particles synthesized by ATRP. Soft Matter 2012;8:4072e82. [34] Dacunha C, Deffieux A, Fontanille M. Synthesis and polymerization of ligninbased macromonomers .3. Radical copolymerization of lignin-based macromonomers with methyl-methacrylate. J Appl Polym Sci 1993;48:819e31. [35] Deoliveira W, Glasser WG. Multiphase materials with lignin. 14. Star-like copolymers with styrene. J Wood Chem Technol 1994;14:119e26. [36] Kim YS, Kadla JF. Preparation of a thermoresponsive lignin-based biomaterial through atom transfer radical polymerization. Biomacromolecules 2010;11: 981e8.
1003
[37] Wang J, Yao K, Korich AL, Li S, Ma S, Ploehn HJ, et al. Combining renewable gum rosin and lignin: towards hydrophobic polymer composites by controlled polymerization. J Polym Sci Pt A: Polym Chem 2011;49:3728e38. [38] Hakem IF, Leech AM, Johnson JD, Donahue SJ, Walker JP, Bockstaller MR. Understanding ligand distributions in modified particle and particle like systems. J Am Chem Soc 2010;132:16593e8. [39] Kramer EJ. Microscopic and molecular fundamentals of crazing. Adv Polym Sci 1983;52-3:1e56. [40] Gent AN, Thomas AG. Effect of molecular weight on the tensile strength of glassy plastics. J Polym Sci Pt A-2: Polym Phys 1972;10:571e3. [41] Kramer EJ. A molecular theory of the fracture toughness of low molecular weight polymers. J Mater Sci 1979;14:1381e8. [42] Feldman D, Banu D, Campanelli J, Zhu H. Blends of vinylic copolymer with plasticized lignin: thermal and mechanical properties. J Appl Polym Sci 2001;81:861e74. [43] Knoll K, Niessner N. Styrolux(þ) and styroflex(þ) e from transparent high impact polystyrene to new thermoplastic elastomers e syntheses, applications and blends with other styrene based polymers. Macromol Symp 1998;132:231e43. [44] Savin DA, Pyun J, Patterson GD, Kowalewski T, Matyjaszewski K. Synthesis and characterization of silica-graft-polystyrene hybrid nanoparticles: effect of constraint on the glass-transition temperature of spherical polymer brushes. J Polym Sci Pol Phys 2002;40:2667e76. [45] Pucciariello R, Villani V, Bonini C, D’Auria M, Vetere T. Physical properties of straw lignin-based polymer blends. Polymer 2004;45:4159e69.
Contents lists available at ScienceDirect
Polymer journal homepage: www.elsevier.com/locate/polymer
A universal route towards thermoplastic lignin composites with improved mechanical properties Shayna L. Hilburg b, Allison N. Elder a, Hoyong Chung a, Rachel L. Ferebee b, Michael R. Bockstaller b, Newell R. Washburn a, c, * a b c
Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, United States Department of Materials Science & Engineering, Carnegie Mellon University, 5000 Forbes Ave, Pittsburgh, PA 15213, United States Department of Biomedical Engineering, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, United States
a r t i c l e i n f o
a b s t r a c t
Article history: Received 11 October 2013 Received in revised form 20 December 2013 Accepted 29 December 2013 Available online 8 January 2014
Nanocomposites based on synthetic polymers grafted from kraft lignin with average particle size of 5 nm were synthesized using atom transfer radical polymerization (ATRP). Lignin macroinitiators were prepared, and polystyrene and poly(methyl methacrylate) were polymerized with target degree of polymerization of 450 resulting in materials having lignin mass fractions of 4.5%, 8.3%, and 22.1% for the poly(methyl methacrylate) samples and 3.2%, 7.1%, and 19.6% for the polystyrene samples. Tensile testing showed a decreased modulus but enhanced toughness of all nanocomposites compared to homopolymers, and the poly(methyl methacrylate)-grafted samples had nearly twice the ultimate elongation as the polystyrene grafts at high graft density. Both types of grafted nanocomposites had toughness values that were greater than 10-times that of the corresponding kraft-lignin/polymer blend system, indicating the potential of ATRP as the basis for the ‘one component’ composite approach towards more sustainable polymeric materials. Dynamical mechanical analysis was used to measure softening temperatures, and both the polystyrene-grafted and poly(methyl methacrylate)-grafted nanocomposites had a peak in the loss modulus that was higher than the corresponding homopolymer, consistent with strong polymer elignin interactions. Lignin grafted with thermoplastic polymers could be an important material based on an inexpensive, renewable feedstock that offers unique mechanical properties compared with many other nanocomposites based on inorganic nanoparticles. Our results indicate that ATRP is well suited for preparing lignin-based thermoplastics and could be the basis for hybrid materials that make effective use of this important renewable resource. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Lignin Nanocomposite ATRP
1. Introduction The integration of natural-sourced polymers with synthetic commodity and engineering plastics is widely recognized to be an important cornerstone in the broader context of developing more sustainable and low-energy polymeric materials [1e8]. Kraft lignin e a term used to describe a range of polydisperse and branched phenolic polymers that are derived from native lignin via the kraft pulping process e has received particular attention due to the relevance of lignin as medium for solar energy storage in plants and its abundance as an important structural component of wood. The high thermal, oxidative, and hydrolytic [9] stability as well as favorable mechanical properties (the Young’s modulus has been
* Corresponding author. Department of Chemistry, Carnegie Mellon University, 4400 Fifth Avenue, Pittsburgh, PA 15213, United States. E-mail address: [email protected] (N.R. Washburn). 0032-3861/$ e see front matter Ó 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.polymer.2013.12.070
estimated to 2 GPa [10]) have motivated interest in using kraft lignin as a filler, for example, for the reinforcement of elastomeric materials [11e18]. However, while the application of lignin in elastomeric as well as thermoset material systems has been widely reported [19], its application as filler in thermoplastic polymeric systems has been a challenge. Among the relevant nonpolar engineering polymer systems, best results were achieved by blending kraft lignin with poly(methyl methacrylate) (PMMA). For example, blends containing 5e20% hydroxypropyl lignin in PMMA were found to have single glass transition temperatures that were around 5 C below that of pure PMMA and a Young’s modulus of 1800e1900 MPa, depending on lignin molecular weight and mass fraction in the blend [20]. This value of Young’s modulus represented a 200 MPa increase over the pure polymer, however, all blends were found to be brittle with a strain-to-fracture less than 2%. The limited effect of kraft lignin was rationalized as a consequence of the poor interfacial binding between filler particles and the matrix that favors particle aggregation and phase separation in
996
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
lignin/polymer blends [21e26]. The development of novel synthetic strategies to facilitate miscible polymer/lignin blends is therefore an important challenge to realize the potential benefits of lignin filler materials for the development of thermoplastic polymer blends. The challenge of facilitating miscible kraft lignin dispersions bears similarity to the broader challenge of understanding the governing parameters that control miscibility in particle/polymer blends [23]. The latter systems have received particular attention in the context of inorganic particle/polymer dispersions. A widely used strategy to compatibilize particle fillers within polymeric hosts is by grafting of polymer chains that are chemically identical to the matrix polymer to the particle matrix [27,28]. In these athermal polymer/particle blends mixing is determined by the subtle interplay of entropic contributions that sensitively depend on the architecture of the polymer-grafted filler (i.e. the density and degree of polymerization of grafted chains as well as the particle size). In general, miscibility is expected only if the degree of polymerization of grafted chains is at least of the same order as the matrix [29]. As recently demonstrated by Ojha et al., [30] the requirement of high molecular grafted chains severely limits the attainable filler concentration and thus the blending of polymergrafted (lignin) fillers with the corresponding matrix polymers cannot be seen as a viable strategy towards lignin/polymer composites with high lignin content and improved property characteristics. An alternative strategy to attain high filler content polymer nanocomposites is based on the concept of ‘one-component’ hybrid materials e nanocomposite materials that are synthesized by the assembly of polymer-grafted particles [31]. Primary advantages of the ‘one-component’ approach are the fine control of the materials microstructure (note that miscibility gaps are absent in onecomponent systems) along with mechanical properties that resemble those of higher molecular weight polymers. Due to their technological relevance as commodity and engineering plastics as well as the accessibility to viable controlled radical polymerization techniques, polystyrene (PS) and PMMA have been studied extensively as model systems in one-component nanocomposites. For example, Choi et al. recently demonstrated that the assembly of PMMA- or polystyrene- (PS) grafted silica particles with suitable architecture enables the fabrication of nanocomposite materials with highly regular microstructure [32]. Interestingly, onecomponent nanocomposites were found to exhibit significantly increased fracture toughness (as compared to the corresponding binary particle/polymer blend systems) e this characteristic was explained as a consequence of the more effective entanglement structure in branched polymeric systems [33]. We note that a particular advantage of the ‘one-component’ approach towards nanocomposite materials is that it is applicable to any type of ‘matrix’ polymer as long as synthetic methodologies exist to facilitate the controlled grafting of the polymer to the particle system. The general concept of one-component nanocomposite materials is illustrated in Scheme 1. The objective of this article is to illustrate the application of ATRP in preparing thermoplastic ‘one-component composites’ that combine a high filling fraction of lignin with improved fracture resistance as compared to the binary polymer/lignin blend systems. The structure of this paper is as follows: In the first part we discuss the conditions that allow for application of atom-transfer radical polymerization to synthesize a set of PMMA- and PS-grafted lignin particle systems with systematic variation of the number and degree of polymerization of surface grafted chains. In the second part we discuss the effect of polymer graft modification on the mechanical properties of lignin-based one-component composites and we demonstrate that even in case of lowest density of polymer
Scheme 1. Comparison of ‘binary’ and ‘one-component’ nanocomposite fabrication schemes.
grafts per lignin ‘particle’ a significant improvement of fracture resistance as compared to ‘classical’ binary composites was observed. Thus, polymer-grafted lignin particles with appropriate architecture could present a novel platform for the fabrication of low-energy thermoplastic commodity polymers with high lignin content and enhanced mechanical properties.
2. Experimental section 2.1. Materials Lignin was purchased from Tokyo Chemical Industry, Co. Methyl methacrylate (MMA), styrene, ethyl 2-bromoisobutyrate (EBiB), CuCl, CuBr, N,N,N0 ,N0 ,N00 -pentamethyldiethylenetriamine (PMDETA), 2,20 -bipyridine (bpy), and pentafluorobenzaldehyde were purchased from SigmaeAldrich. Styrene and MMA monomers were run through a basic alumina column before use to remove inhibitor. All other chemicals were used as received.
2.2. ATRP of styrene First, styrene, dimethylformamide (DMF), PMDETA, and EBiB were dried over molecular sieves and degassed under nitrogen for 30 min. Then CuBr (14.8 mg, 0.103 mmol) was added to a 50 mL Schlenk flask, which was subsequently degassed by three vacuume nitrogen cycles. Next, 4.5 mL DMF, 21.5 mL (0.103 mmol) PMDETA, 5.31 mL (46.3 mmol) styrene, and 15.1 mL (0.103 mmol) EBiB were added to the flask and the reaction was carried out at 100 C for 7 days. The percent conversion was determined to be 52% using 1H NMR by comparing the disappearance of acrylate peaks to DMF internal standard. Copper was removed by dilution with THF and filtration through neutral alumina. The polymer was precipitated into 100 mL of hexane and dried in vacuo at room temperature. The resulting polymer (1.2401 g, 51.4% yield) was a white brittle solid.
2.3. ATRP of methyl methacrylate ATRP of MMA was conducted in the same manner as described above. First, MMA, DMF, PMDETA, and EBiB were bubbled under nitrogen for 30 min and 17 mg (0.119 mmol) of CuBr was added to a Schlenk flask with three vacuumenitrogen cycles. MMA (5.03 mL, 47.2 mmol), 1.2 mL DMF, PMDETA (21.9 mL, 0.105 mmol), and EBiB (15.4 mL, 0.105 mmol) were added and the reaction was carried out at 70 C for 24 h. Percent conversion according 1H NMR was 62%. Again, copper was removed by dilution with THF and filtration through neutral alumina. A total of 1.9543 g (66.7% yield) was obtained after precipitation into hexane and drying in vacuo.
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
997 OH
OH
Lignin
O HO
Lignin
O Br
Br
HO
O Pyridine, THF
H or Lignin
H3CO
R.T. 24 hours
H or Lignin
H3CO
O
OH
O
Br
Scheme 2. Synthesis of lignin macroinitiator.
2.4. Lignin acidification
2.6. ATRP of styrene-grafted lignin samples
Lignin (5.00 g) was suspended in 75 mL of 2 M HCl with stirring for 15 min. The mixture was vacuum filtered and the acidification/ filtration process was repeated. During the second filtration, the solid lignin was rinsed with water and then allowed to air-dry overnight. Finally, the solid powder was dried under vacuum for 5 h at 35 C. The total amount obtained was 3.30 g giving a 66% yield.
For polymer graft synthesis using the macroinitiator with highest concentration of initiators, 0.3082 g lignin macroinitiator, 21.0 mg CuCl (0.212 mmol), and 58.2 mg bpy (0.373 mmol) were added to the Schlenk flask and degassed using three vacuumenitrogen cycles. DMF and styrene were bubbled under nitrogen for 30 min then DMF (8.5 mL) and styrene (10.4 mL, 90.8 mmol) were added to the flask. The reaction was carried out at 100 C for up to 14 days, which provided the highest reproducibility. However, most reactions can be terminated after a few days. The percent conversion from 1H NMR was determined to be 97.2%. The polymer was then diluted with THF, precipitated into hexane, and dried under vacuum at room temperature. Additional styrene-grafted polymers were synthesized using the same procedure. These polymers consisted of styrene grafted from lignin with lignin fractions of 4.5%, 8.3%, and 22.1%, and the amount of monomer, ligand, and catalyst were scaled in accordance with the moles of initiator. Scheme 3 presents a summary of the polymerization conditions.
2.5. Lignin macroinitiator synthesis Acidified lignin (1.06 g) was dissolved in 20 mL of pyridine in a 200 mL roundbottom flask. The lignin solution was bubbled under nitrogen for 30 min, followed by the addition of 10 mL of deoxygenated THF. To prepare samples with the highest polymer content (4.5% lignin), 186 mL (0.261 g, 1.14 mmol) of a-bromoisobutyryl bromide was added dropwise. The flask was then allowed to stir for 24 h at room temperature. After 24 h, the lignin was precipitated into 400 mL diethyl ether. The ether was decanted off and the resulting solid was dried at 35 C under vacuum. The total amount of initiator obtained was 1.17 g (88% yield). The number of initiator sites per mass of lignin was determined by 1H NMR in DMSO with 10 mL of pentafluorobenzaldehyde (PFB) as an internal standard. 1H NMR (300 MHz) d ¼ 2.0 ppm (s, 6 H, conjugated initiator), d ¼ 10.19 ppm (s, 1 H, PFB). Additional lignin initiators with lower concentrations of initiator were synthesized using the same procedure, but scaling down the a-bromoisobutyryl bromide accordingly. A summary of the conditions for lignin macroinitiator synthesis is shown in Scheme 2.
2.7. ATRP of methyl methacrylate-grafted lignin samples For polymer graft synthesis using the macroinitiator with the highest polymer:lignin ratio, 0.1915 g initiator (ca 0.126 mmol), 18.8 mg CuBr (0.131 mmol), and 26.3 mL PMDETA (0.126 mmol) were added to the Schlenk flask and degassed using three vacuume nitrogen cycles. DMF and MMA were bubbled under nitrogen for 30 min, then DMF (7.5 mL) and MMA (6.0 mL, 56.3 mmol) were added to the flask. The reaction was carried out at 70 C for 24 h.
OH
OH
Lignin
O
Lignin
O
HO
HO
H or Lignin
H3CO O
H or Lignin
CuCl, bpy, DMF 100°C
O
O
O
Br
n
Scheme 3. Synthesis of styrene-grafted lignin.
998
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003 OH
OH
Lignin
O
Lignin
O
O
HO
HO
O
H or Lignin
H3CO
H or Lignin
PMDETA, CuBr, DMF 70°C
O
O O
O
Br
O n
O
Scheme 4. Reaction conditions for synthesis of MMA-grafted lignin.
The percent conversion from 1H NMR was determined to be 72.3%. The polymer was then diluted with THF, precipitated into hexane, and dried under vacuum at room temperature. Additional MMA-grafted polymers with lower polymer:lignin ratios were synthesized using the same procedure. To ensure high conversion, reactions were run up to 8 days, but tuning reaction conditions can increase rates. Scheme 4 presents a summary of the reaction conditions for MMA-grafted lignin. 2.8. NMR NMR measurements were carried out using a Bruker Advance 300 MHz NMR Instrument in CDCl3 or deuterated dimethyl sulfoxide. 2.9. Tensile testing
(a)
Tensile stress-strain tests were performed on an Instron tensile test machine using pneumatic grips and operated at a crosshead speed of 0.15 mm/min. Specimen for tensile test evaluation were prepared by first hotpressing of samples at 120 C and subsequent cutting of strips with test region dimensions of 20 mm 5 mm 1.5 mm. 2.10. Dynamic mechanical analysis (DMA) Disks of approximately 8 mm in diameter and 2 mm height were prepared from each sample. Polymer was pressed into a cylindrical mold at 120 C then annealed at 100 C under vacuum for at least 48 h. DMA was performed on a TA Instruments RSA-G2 using 15 mm disk compression clamps. Samples were equilibrated at 150 C before oscillating at 0.25% strain at 1 Hz while decreasing temperature at 1 C/min.
(b)
3. Results and discussion 3.1. Chemical and structural analysis Representative NMR data for unmodified lignin, lignin modified with the ATRP initiator, and MMA-grafted lignin (4.5% lignin) are shown in Fig. 1(a)e(c), respectively. Due to lignin being present in a very small mass fraction (<5%), the lignin peaks for the grafted material are difficult to see. Pentafluorobenzaldehyde was added to the sample shown in Fig. 1(b) to allow for quantitative measurement of the moles of initiator per gram of lignin. Additional peaks in the grafted material shown in Fig. 1(c) appear at 7.9, 2.7, and 2.8 ppm and are due to residual DMF from the polymerization
(c) Fig. 1. 1H NMR of lignin (a), lignin macroinitator (b), and lignin-g-PMMA (c).
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
999
Table 1 Compositions of samples reported in this study. Sample
Polymer graft
Average DP
Mass % lignin
lignin22.1PMMA401 lignin8.3PMMA388 lignin4.5PMMA323 lignin19.6PS449 lignin7.1PS444 lignin3.2PS437
PMMA PMMA PMMA PS PS PS
401 388 323 449 444 437
22.1 8.3 4.5 19.6 7.1 3.2
reaction and at 1.4 due to hexane. Based on quantification of the spectra, the initiator site concentrations for the three macroinitiators were measured to be 80, 280, and 640 mmol/(g lignin). Based on a GPC-derived lignin molecular weight of 25,000 g/mol, this corresponds to grafting densities of approximately 2, 7, and 16 per lignin molecule. However, initiator site density will be reported based on mass of lignin instead of per lignin particle because of the high level of heterogeneity in lignin. While it is a three-dimensional polymer, in using it as a macroinitiator we treat it similarly to inorganic nanoparticles, which are represented on a “per mass” basis instead of “per molecule” in one-component nanocomposites. The solubility of the products indicates polymerization from the lignin surface has been performed. Attachment of initiators offers high monomer conversion and essentially quantitative grafting efficiency compared with other methods, such as radiationinduced polymerization or reacting lignin with methacryloyl chloride, [34] in providing a means for continuously varying composition without forming crosslinked products. Reactive blending of lignin generally results in less than 100% grafting efficiency even using isocyanate-terminated polymers, suggesting that grafting-from chemistries provide a more efficient means for surface modification [35]. Indeed, Kim and Kadla demonstrated the utility of ATRP in polymerization of N-isopropylacrylamide from lignin macroinitiators to prepare thermal responsive materials [36]. A similar approach was taken by Tang and co-workers in the preparation of hydrophobic polymer-grafted lignin by ATRP of rosin polymers from a lignin macroinitiator [37]. The current research extends these approaches to making high-strength one-component composites, demonstrating ATRP offers significant advantages over other approaches taken to synthesize lignin grafted with thermoplastic polymers. The compositions of the six polymer-grafted lignin nanocomposites synthesized in this work are summarized in Table 1. Samples are identified as ligninxxPOLYMERyyy where xx is the mass fraction of lignin in the sample, POLYMER is PMMA or PS, and yyy is the average degree of graft polymerization based on monomer conversion. All samples were dark brown in color and dissolved readily in pyridine or DMF. 3.2. Structural characterization In Fig. 2 are shown the (intensity-averaged) distribution of hydrodynamic radii (RH) for the PMMA homopolymer, kraft lignin, and lignin4.5PMMA323 that were obtained by DLS analysis in DMF solution. The peak in the size distribution of lignin was 5 nm, PMMA homopolymer was 7 nm, while the value for the PMMAgrafted lignin was 28 nm. There is an aggregation in the dissolved copolymers, evidenced by a peak at larger particle size, but not to a large extent. Note also that the shape of the distribution of hydrodynamic radii is consistent with the expected heterogeneity of the polymer-conjugated product [38] e in particular the absence of a shoulder at small RH suggests that ATRP initiated cleanly from the lignin macroinitiator.
Fig. 2. Dynamic light scattering data from DMF solutions of kraft lignin (top), PMMA homopolymer (middle), and lignin4.5PMMA323 (bottom).
The conclusion of successful graft modification is also confirmed by transmission electron microscopy of solution-cast films of polymer-grafted lignin particles. Films were deposited from a 1 mg/ mL pyridine solution onto TEM grids and stained with OsO4 (selective for lignin) to provide contrast between lignin and PMMA. Fig. 3a depicts a representative TEM image of a thin film of lignin4.5PMMA323 revealing a uniform dispersion of lignin particles (dark) in PMMA matrix. Interestingly, the average nearest-neighbor particle distance hdi w 15 nm that can be determined from the micrograph (see inset of Fig. 3a) is consistent with the presence of PMMA grafts with degree of polymerization N ¼ 323 for which the radius of gyration of surface-grafted chains can be estimated to Rg,PMMA323 w 7 nm. To assess the implication of polymer graft modification on the mechanical properties of lignin composites, reference systems comprised of binary blends of lignin dispersed within the respective host matrix were prepared. For example, Fig. 3b depicts a representative electron micrograph of a binary lignin/PMMA blend with equal composition (i.e. lignin content) as the one-component lignin composite shown in Fig. 3a. The figure reveals a dispersed particle morphology in which the average diameter of particles is found to be of order 10 nm, similar to what was measured using DLS for individual lignin particles. Images of PMMA-grafted materials with lower graft densities appeared to have greater direct ligninelignin interactions, consistent with the mass fraction of lignin increasing from 4.5% to 8.3% and 22.1%, although the higher staining levels made it difficult to clearly observe the morphology (images not shown). The miscibility of lignin in PMMA may be understood as a consequence of the compatibilization effect that is imparted by hydroxyl functionalities on the surface of lignin that allow for hydrogen-bond formation with the PMMA host. Note that both systems e the one-component lignin4.5PMMA323 as well as the binary lignin/PMMA blend e are equal in morphology and composition and therefore are ideally suited to identify the relevance of molecular connectivity on the properties of the resulting composite material. We also note that lignin was found to be incompatible in PS host matrices, for which aggregated morphologies were observed (results not shown here). 3.3. Mechanical and thermal characterization The mechanical properties of the PMMA-lignin nanocomposites were characterized using tensile testing at room temperature.
1000
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
the effect of lignin content. Since PMMA and PS nanocomposites with lowest grafting densities had similar stressestrain curves, with ultimate elongations of 20% and tensile strengths of 5 MPa, we conclude that at higher lignin content the mechanical properties of the one-component nanocomposites were primarily determined by the lignin component. Interestingly, one-component lignin-g-PMMA systems were found to significantly exceed the fracture toughness of the corresponding binary blend systems. Lignin4.5PMMA323 had a toughness of 9.6 MJ/m3, which was more than 10-times greater than that of the PMMA homopolymer or the binary lignin-PMMA blend reference system. The improved toughness and ductility of onecomponent systems is consistent with recent reports on the mechanical properties of one-component composites based on PMMA-grafted silica particle systems that also suggest an enhancement of fracture toughness in one-component composite systems as compared to the reference binary blend systems [33]. In analogy to the inorganic homologues we interpret the increase of fracture toughness to be a consequence of the more efficient connectivity among entanglement points in the star-polymer-like lignin4.5PMMA323. The latter should support the interaction between entanglement points during craze formation and hence raise the energy dissipation during fracture. The non-continuous trend of modulus and toughness that is observed in both polymer-grafted lignin one-component composite materials with increasing number of polymer is surprising and counterintuitive, however, since in both cases the best performance was observed for
Fig. 3. Representative bright-field TEM images of lignin4.5PMMA323 composite materials. Panel a: lignin4.5PMMA323 one-composite revealing uniform particle dispersion. Scale bar is 200 nm. Inset shows distribution of next-nearest neighbor distances. Average distance is determined as hdi z 25 nm consistent with the presence of PMMA grafts with N ¼ 323. Panel b: binary lignin/PMMA blend system with equal volume fraction lignin as system shown in (a). Micrograph reveals uniform lignin-particle dispersion in PMMA. The scale bar is 500 nm. Inset shows magnified view of highlighted area depicting individual lignin particles dispersed within PMMA matrix. Scale bar is 20 nm.
Stressestrain curves were recorded for each polymer-grafted sample and compared with homopolymers of PMMA and PS having similar average molecular weight as the polymer grafts as well as the 1:7 binary blend reference system of lignin and PMMA homopolymer. Representative curves are shown in Fig. 4(a) and (b) for the grafted nanocomposites. The PMMA samples had ultimate elongation and toughness values that were equal to or greater than the values for the PS samples. This is consistent with previous results that have related the toughness of PMMA to the presence of a b-relaxation process associated with the motion of side chains [39]. Similarly, lignin4.5PMMA323 was found to exhibit a toughness value about twice the value of lignin3.2PS437. This likely reflects both intrinsic differences in the mechanical properties of PMMA and PS as well as
Fig. 4. Stress versus strain curves for (a) lignin-g-PMMA and (b) lignin-g-PS.
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
the lowest and highest concentrations of polymer grafts, respectively. We note that similar trends are observed for both material systems and thus we consider the result to be reliable. We hypothesize that the varying effect of polymer graft modification is a consequence of the competing influence of distinct governing parameters (such as the lignin volume fraction, the degree of polymerization and the conformation of grafted chains) on the mechanical properties of polymer-grafted lignin composite materials. Note that homopolymers and ligninehomopolymer blends were found to exhibit similar mechanical properties for PS and PMMA. The PS homopolymer had a modulus of 2700 MPa while the PMMA homopolymer was only 280 MPa, but the toughness values were significantly lower than that of high molecular weight analogues [40,41]. Furthermore, while the ultimate elongation of the PS was 5%, similar to what would be expected from a brittle polymer, that of PMMA was 15%, suggesting the latter is less brittle at lower molecular weights [39]. Solvent blending of lignin with each homopolymer resulted in modest increases in both modulus and toughness, although the magnitude of the increases was lower than what previously reported values [24,26,42,43]. Other researchers have demonstrated significant increases in modulus in particular, and the lack of increase may be due to the relatively low molecular weight of homopolymers used in the present study. Dynamical mechanical analysis (DMA) was performed on the samples to determine the softening temperatures of the samples, which can be associated with a glass transition. However, preliminary differential scanning calorimetry measurements on the grafted materials only showed broad thermal transitions (data not shown), suggesting these materials may not be true glass formers. Further study is necessary to clarify the physical nature of the softening transitions observed in DMA, but the peak in the loss modulus is reported as Tg for these purposes. Storage and loss moduli were taken from each of the six polymer samples, and representative plots for each sample are shown in Fig. 5. Samples were equilibrated at 150 C for 10 min, and data were collected as the samples cooled were cooled at 1 C/min. The storage modulus rises sharply as the samples cool through Tg, and the loss modulus exhibits the expected peak associated with this thermal transition. The trend in storage modulus for both PMMAand PS-grafted materials reflects that observed in Young’s modulus. Interestingly, Fig. 5 reveals that in the limit of small number of grafted chains the relaxation appears broadened both in case of PMMA and PS grafted lignin. This is consistent with the observed increase in toughness in case of a small number of polymer grafts and indicates a complex relationship between the architecture of grafted lignin and the mechanical properties of one-component lignin composites. The pure PMMA and PS samples had Tg values of 65 C and 80 C, respectively, while the 1:7 lignin blends had softening temperatures of 92 C and 95 C, respectively. These curves are omitted in Fig. 5 for clarity but had similar shapes as the grafted materials. The significant increase in Tg for the lignin-PMMA blend compared to the PMMA homopolymer suggests strong interactions between the nanoparticle and polymer matrix. The increase in Tg across all compositions is consistent with other measurements of polymer-grafted nanoparticles and is generally expected due to the constraints imparted on chain relaxation by bonding of one chain end to the inorganic particle. For example, silica-PS nanocomposites had Tg values that averaged 13 C greater than the corresponding homopolymer, which was attributed to chain confinement on the nanoparticle surface [44]. However, the 27 C increase is greater than that observed by Ciemniecki and Glasser, [20] who observed little or no difference in Tg when blending hydroxypropyl lignin into PMMA using tetrahydrofuran compatibilization, but this too may be due to chemically
1001
Fig. 5. Storage (black) and loss (gray) moduli for (a) lignin-g-PMMA samples and (b) lignin-g-PS samples with varying numbers of initiator sites per lignin.
functionalized lignin used in their study as well as differences in solvent processing. Table 2 summarizes the material parameters that were obtained through data fitting of tensile testing and DMA characterization. The toughening observed in these grafted one-component nanocomposites contrasts with the behavior observed in many ligninepolymer blends. Blends of up to 40% hydroxypropyl lignin in PMMA, investigated by Ciemniecki and Glasser, resulted in materials with a phase-separated morphology but one that displayed a single Tg that was insensitive to the lignin content [20]. The tensile strength of these blends varied from 50 MPa to 87 MPa depending on the lignin formulation and blending conditions, but the ultimate elongation values decreased relative to the unfilled polymer upon blending with even 5% lignin. In a separate study by Pucciariello et al., blends containing 10% straw lignin in PS having Mw of 280,000 showed a tensile strength of 12.0 MPa and an ultimate elongation of 11.8%, although the latter parameter decreases rapidly upon increasing the lignin content, suggesting this approach is not capable of increasing the amount of a renewable feedstock in PS [45]. One formulation with significantly different mechanical
1002
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003
Table 2 Young’s modulus, toughness, ultimate elongation, and glass transition temperature values for PMMA and PS materials. Polymer
Young’s modulus (MPa)
lignin22.1PMMA401 lignin8.3PMMA388 lignin4.5PMMA323 PMMA (25 kDa) lignin/PMMA blend (1:7) lignin19.6PS449 lignin7.1PS444 lignin3.2PS437 PS (25 kDa) lignin/PS blend (1:7)
56 46 55 100 121 62 34 81 3100 2700
1 26 3 14 23 13 11 25 132 210
Toughness (MJ/m3) 1.0 2.7 3.7 0.15 0.24 0.37 0.28 0.52 0.012 0.014
0.3 1.4 0.1 0.1 0.1 0.25 0.05 0.24 0.002 0.002
Ultimate elongation (%) 23 58 52 12 13 12 15 14 3 4
4 26 4 5 3 7 1 7 1 1
Tg ( C)
77 75 82 65 92 79 103 96 80 95
properties was a blend containing 85% kraft lignin in a poly(vinyl acetate) matrix [25]. By carefully controlling the molecular weight of the lignin through a pH-dependent precipitation process, Sarkanen and co-workers were able to produce a blend having high lignin content that displayed an ultimate elongation of greater than 60% and a tensile strength of nearly 10 MPa [25]. However, the thermodynamic compatibility of kraft lignin and poly(vinyl acetate) was an important design principle, and poly(vinyl acetate) is not commonly used as a structural material in part due to its susceptibility to hydrolysis and conversion to poly(vinyl alcohol). The pronounced increase in fracture toughness that is realized in polymer-grafted lignin composites, while largely retaining the materials stiffness, renders the ‘one-component’ approach an appealing strategy in the fabrication of lignin-based thermoplastic engineering polymers.
4. Conclusion ATRP is well suited for preparing lignin-based nanocomposites with grafted thermoplastic polymers. Characterization using DLS confirmed that polymerization proceeds cleanly from the lignin macroinitiators and results in products defined by a lignin core with a polymer corona. Electron microscopy images of the ligning-PMMA sample having lignin mass fraction of 4.5% showed that the microstructure was based on a lignin core in a continuous matrix of PMMA. Tensile testing and DMA were used to characterize the mechanical properties and thermodynamic properties of lignin-g-PS and lignin-g-PMMA nanocomposites. The increase of Tg as compared to the homopolymer systems is consistent with the restriction of grafted polymer chains and indicates that interactions between lignin and both polymers is strong, but particularly in the lignin-g-PMMA samples, where the highest toughness values were observed. However, the highest moduli were observed for the compositions with the highest lignin content, suggesting the properties become less dependent on the specific polymer and more dependent on the intrinsic lignin properties. One-component composites based on polymer-grafted lignin were found to exhibit significantly enhanced fracture toughness as compared to binary lignin/polymer blends while retaining the stiffness characteristics of polymer glasses. The fortuitous combination of mechanical properties along with the fine control of the composite microstructure should render polymer-grafted lignin composites a platform for developing advanced materials based on renewable resources. This is particularly true since the current progress in controlled radical polymerization suggest the application of the approach to a wide range of lignin-based nanocomposites.
Acknowledgments Dr. Ludwig Leibler is gratefully acknowledged for helpful discussions. NMR instrumentation at CMU was partially supported by NSF (CHE-0130903 and CHE-1039870). R.L.F. and M.R.B. acknowledge support through the NSF-IGERT program via NSF-0966227. A.N.E. gratefully acknowledges the Richard King Mellon Presidential Fellows program for support.
References [1] Bataille P, Ricard L, Sapieha S. Effects of cellulose fibers in polypropylene composites. Polym Compos 1989;10:103e8. [2] Bledzki AK, Gassan J. Composites reinforced with cellulose based fibres. Prog Polym Sci 1999;24:221e74. [3] Chung H, Washburn NR. Chemistry of lignin-based materials. Green Mater 2012;1:137e60. [4] Feldman D. Lignin polymer systems and some applications. Prog Polym Sci 1986;12:271. [5] Li Y, Sarkanen S. Alkylated kraft lignin-based thermoplastic blends with aliphatic polyesters. Macromolecules 2002;35:9707e15. [6] Lundquist L. Life cycle engineering of plastics: technology, economy, and the environment. Oxford; New York: Elsevier; 2000. [7] Ratto JA, Stenhouse PJ, Auerbach M, Mitchell J, Farrell R. Processing, performance and biodegradability of a thermoplastic aliphatic polyester/starch system. Polymer 1999;40:6777e88. [8] Rodriguez-Gonzalez FJ, Ramsay BA, Favis BD. High performance LDPE/thermoplastic starch blends: a sustainable alternative to pure polyethylene. Polymer 2003;44:1517e26. [9] Sun Y, Cheng JY. Hydrolysis of lignocellulosic materials for ethanol production: a review. Bioresour Technol 2002;83:1e11. [10] Cousins WJ. Elastic-modulus of lignin as related to moisture-content. Wood Sci Technol 1976;10:9e17. [11] Cateto CA, Barreiro MF, Rodrigues AE, Brochier-Solan MC, Thielemans W, Belgacem MN. Lignins as macromonomers for polyurethane synthesis: a comparative study on hydroxyl group determination. J Appl Polym Sci 2008;109:3008e17. [12] Chung HY, Washburn NR. Improved lignin polyurethane properties with Lewis acid treatment. Acs Appl Mater Inter 2012;4:2840e6. [13] Hirose S, Hatakeyama T, Hatakeyama H. Synthesis and thermal properties of epoxy resins from ester-carboxylic acid derivative of alcoholysis lignin. Macromol Symp 2003;197:157e69. [14] Morck R, Reimann A, Kringstad KP. Elastomeric polyurethanes from a kraft lignin polyethylene-glycol toluene diisocyanate system. Acs Sym Ser 1989;397:390e401. [15] Nakano J, Izuta Y, Orita T, Hatakeyama H, Kobashigawa K, Teruya K, et al. Thermal and mechanical properties of polyurethanes derived from fractionated kraft lignin. Sen-I Gakkaishi 1997;53:416e22. [16] Reimann A, Morck R, Yoshida H, Hatakeyama H, Kringstad KP. Kraft lignin in polyurethanes. 3. Effects of the molecular-weight of Peg on the properties of polyurethanes from a kraft lignin Peg Mdi system. J Appl Polym Sci 1990;41: 39e50. [17] Thring RW, Vanderlaan MN, Griffin SL. Polyurethanes from Alcell (R) lignin. Biomass Bioenergy 1997;13:125e32. [18] Yoshida H, Morck R, Kringstad KP, Hatakeyama H. Kraft lignin in polyurethanes. 1. Mechanical-properties of polyurethanes from a kraft lignin polyether triol polymeric Mdi system. J Appl Polym Sci 1987;34:1187e98. [19] Lora JH, Glasser WG. Recent industrial applications of lignin: a sustainable alternative to nonrenewable materials. J Polym Environ 2002;10:39e48. [20] Ciemniecki SL, Glasser WG. Multiphase materials with lignin. 1. Blends of hydroxypropyl lignin with Poly(Methyl methacrylate). Polymer 1988;29: 1021e9. [21] Chen F, Dai HH, Dong XL, Yang JT, Zhong MQ. Physical properties of ligninbased polypropylene blends. Polym Compos 2011;32:1019e25. [22] Feldman D, Banu D, Manley RSJ, Zhu H. Highly filled blends of a vinylic copolymer with plasticized lignin: thermal and mechanical properties. J Appl Polym Sci 2003;89:2000e10. [23] Kadla JF, Kubo S. Lignin-based polymer blends: analysis of intermolecular interactions in lignin-synthetic polymer blends. Compos Part A-Appl S 2004;35:395e400. [24] Kubo S, Kadla JF. The formation of strong intermolecular interactions in immiscible blends of poly(vinyl alcohol) (PVA) and lignin. Biomacromolecules 2003;4:561e7. [25] Li Y, Mlynar J, Sarkanen S. The first 85% kraft lignin-based thermoplastics. J Polym Sci Pol Phys 1997;35:1899e910. [26] Wang JS, Manley RS, Feldman D. Synthetic-polymer lignin copolymers and blends. Prog Polym Sci 1992;17:611e46. [27] Rong MZ, Zhang MQ, Zheng YX, Zeng HM, Walter R, Friedrich K. Structureproperty relationships of irradiation grafted nano-inorganic particle filled polypropylene composites. Polymer 2001;42:167e83.
S.L. Hilburg et al. / Polymer 55 (2014) 995e1003 [28] Balazs AC, Emrick T, Russell TP. Nanoparticle polymer composites: where two small worlds meet. Science 2006;314:1107e10. [29] Green PF. The structure of chain end-grafted nanoparticle/homopolymer nanocomposites. Soft Matter 2011;7:7914e26. [30] Ojha S, Dang A, Hui CM, Mahoney C, Matyjaszewski K, Bockstaller MR. Strategies for the synthesis of thermoplastic polymer nanocomposite materials with high inorganic filling fraction. Langmuir: ACS J Surfaces Colloids 2013;29:8989e96. [31] von Werne T, Patten TE. Preparation of structurally well-defined polymernanoparticle hybrids with controlled/living radical polymerizations. J Am Chem Soc 1999;121:7409e10. [32] Choi J, Dong H, Matyjaszewski K, Bockstaller MR. Flexible particle array structures by controlling polymer graft architecture. J Am Chem Soc 2010;132:12537e9. [33] Choi J, Hui CM, Pietrasik J, Dong HC, Matyjaszewski K, Bockstaller MR. Toughening fragile matter: mechanical properties of particle solids assembled from polymer-grafted hybrid particles synthesized by ATRP. Soft Matter 2012;8:4072e82. [34] Dacunha C, Deffieux A, Fontanille M. Synthesis and polymerization of ligninbased macromonomers .3. Radical copolymerization of lignin-based macromonomers with methyl-methacrylate. J Appl Polym Sci 1993;48:819e31. [35] Deoliveira W, Glasser WG. Multiphase materials with lignin. 14. Star-like copolymers with styrene. J Wood Chem Technol 1994;14:119e26. [36] Kim YS, Kadla JF. Preparation of a thermoresponsive lignin-based biomaterial through atom transfer radical polymerization. Biomacromolecules 2010;11: 981e8.
1003
[37] Wang J, Yao K, Korich AL, Li S, Ma S, Ploehn HJ, et al. Combining renewable gum rosin and lignin: towards hydrophobic polymer composites by controlled polymerization. J Polym Sci Pt A: Polym Chem 2011;49:3728e38. [38] Hakem IF, Leech AM, Johnson JD, Donahue SJ, Walker JP, Bockstaller MR. Understanding ligand distributions in modified particle and particle like systems. J Am Chem Soc 2010;132:16593e8. [39] Kramer EJ. Microscopic and molecular fundamentals of crazing. Adv Polym Sci 1983;52-3:1e56. [40] Gent AN, Thomas AG. Effect of molecular weight on the tensile strength of glassy plastics. J Polym Sci Pt A-2: Polym Phys 1972;10:571e3. [41] Kramer EJ. A molecular theory of the fracture toughness of low molecular weight polymers. J Mater Sci 1979;14:1381e8. [42] Feldman D, Banu D, Campanelli J, Zhu H. Blends of vinylic copolymer with plasticized lignin: thermal and mechanical properties. J Appl Polym Sci 2001;81:861e74. [43] Knoll K, Niessner N. Styrolux(þ) and styroflex(þ) e from transparent high impact polystyrene to new thermoplastic elastomers e syntheses, applications and blends with other styrene based polymers. Macromol Symp 1998;132:231e43. [44] Savin DA, Pyun J, Patterson GD, Kowalewski T, Matyjaszewski K. Synthesis and characterization of silica-graft-polystyrene hybrid nanoparticles: effect of constraint on the glass-transition temperature of spherical polymer brushes. J Polym Sci Pol Phys 2002;40:2667e76. [45] Pucciariello R, Villani V, Bonini C, D’Auria M, Vetere T. Physical properties of straw lignin-based polymer blends. Polymer 2004;45:4159e69.