Ductile nanocellulose-based films with high stretchability and tear resistance

European Polymer Journal 69 (2015) 328–340

Contents lists available at ScienceDirect

European Polymer Journal journal homepage: www.elsevier.com/locate/europolj

Macromolecular Nanotechnology

Ductile nanocellulose-based films with high stretchability and tear resistance Steven Spoljaric, Arto Salminen, Nguyen Dang Luong, Jukka Seppälä ⇑ Polymer Technology, Department of Biotechnology and Chemical Technology, Aalto University School of Chemical Technology, P.O. Box 16100, 00076 Aalto, Finland

a r t i c l e

i n f o

Article history: Received 18 February 2015 Received in revised form 5 June 2015 Accepted 16 June 2015 Available online 17 June 2015 Keywords: Nanofibrillated cellulose Hydroxyethylcellulose Ductility Tear strength

a b s t r a c t Hybrid composite films of nanofibrillated cellulose (NFC), cationic hydroxyethylcellulose (CHEC) and glycerol were prepared via solution blending. NFC adopted a fibrillar structure, being coated with the CHEC moieties. Ionic interactions between NFC and CHEC were present, as confirmed by FTIR, leading to enhanced strength, moduli and work-to-break values. Simultaneously, segments of non-cationic HEC worked in tandem with glycerol to lubricate the cellulose nanofibrils, yielding maximum strain-at-break values of 141%. The NFC–CHEC– glycerol films were subsequently coated with a 4.5–6 lm layer of poly(propylene carbonate) (PPC), microscopy confirming effective coating of NFC at the composite-coating layer boundary. The hydrophobic nature of PPC enhanced water stability for acute time periods and increased contact angle values. Film strength and toughness was also enhanced by PPC, indicative of its reinforcing ability. The hybrid composites highlight the ability to prepare extremely ductile cellulose-based films, with the aforementioned materials being amongst the most ductile cellulose-based and/or cellulose-majority films documented. Ó 2015 Published by Elsevier Ltd.

1. Introduction Lignocellulosic materials have rapidly become a popular material and research field, driven by their many benefits including renewability, sustainability, recyclability, low cost and biodegradability [1]. Within this class of natural materials, nanofibrillated cellulose (NFC) is among the most studied and most promising. NFC is a nanomaterial derived from wood, prepared from the fragmentation of cellulose within wood pulp. The process yields cellulose nanofibrils with a high aspect ratio (4–20 nm wide, 500–2000 nm in length), high stiffness and mechanical properties and a low percolation threshold. This combination of favourable characteristics and nanoscale dimensions has encouraged the use of NFC-reinforcement in polymer nanocomposites, with composites displaying enhanced mechanical [2] and thermal [3] behaviour compared to materials filled with traditional wood-pulp. The nanoscale dimensions of the fibrils allow property enhancement to occur without compromising optical properties. The interest in NFC as a composite component is further fuelled by the ongoing focus on environmentally friendly and ‘green’ composites and materials. Despite its many advantages, the utilisation of NFC presents several drawbacks and limitations. Most notably, NFC films are extremely brittle and possess poor ductility and tear strength. Films prepared from pure NFC seldom reach strain at break values of 10% [4,5]. The most common methods of enhancing ductility within cellulose films involve the incorporation of plasticisers such as glycerol, sorbitol and poly(ethylene glycol) [6]. More recently, processing techniques including multi-layer formation and over-pressure filtration and hot pressing [7] have been utilised to increase elongation. Sehaqui ⇑ Corresponding author. E-mail address: jukka.seppala@aalto.fi (J. Seppälä). http://dx.doi.org/10.1016/j.eurpolymj.2015.06.019 0014-3057/Ó 2015 Published by Elsevier Ltd.

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

329

et al. [8,9] demonstrated how the ductility of NFC can be increased by blending with hydroxyethylcellulose (HEC), reaching maximum strain values of 55%. Furthermore, blending with HEC has increased ductility in other naturally-derived materials, including bacterial cellulose [10] and chitosan [11]. Thus, there is great potential in determining the limits of ductility and elasticity enhancement in natural composites and blends via the incorporation of HEC. As is quite often the case with polymer composites, increased ductility and elasticity comes at the expense of strength and toughness. In order to retain some of the initial mechanical characteristics, chemical modification of the cellulosic materials can be performed to facilitate superior interaction and adhesion. NFC has the advantage of possessing a slight negative charge. Therefore, its polyelectrolytic nature can be taken advantage of via the introduction of cationic groups. HEC can be readily cationised using simple and safe methods [12,13]. Compounding cationised-HEC (CHEC) with NFC can lead to enhanced strength and toughness due to ionic complexation between both cellulosic components. Since several hydroxyl groups that constitute HEC would not be cationised, segments of HEC can also interact with NFC via hydrogen bonding and coat the nanofibrils [9,10]. The primary mechanism influencing film ductility arises from this interaction, as HEC-coated NFC fibrils experience slippage and alignment instead of fracturing under applied load. Thus, a synergy between the strength obtained from ionic bonding and the lubrication effect of HEC coating the cellulose nanofibrils is necessary. To the best of the authors’ knowledge, the bulk of research concerning HEC-cellulose blends focus on physical interactions (hydrogen bonding, etc.) between the components rather than polyelectrolytic complexation. One disadvantage of utilising electrostatic interactions is that potential enhancements in strength and modulus may be at the expense of film ductility. A possible method of overcoming this is to incorporate a compound, such as glycerol, into the cellulosic composite. Glycerol may effectively lubricate NFC and CHEC segments, allowing for slippage during the application of load and subsequently enhancing material ductility. This provides prospects to determine the limits of ductile and elastic behaviour within composites containing maximal cellulose loadings. Despite its potential to enhance ductility within nanocellulose films, glycerol possesses a tendency of migration and leaching from cellulose films. Due to its polar nature, glycerol readily adsorbs to water molecules. In order to prevent glycerol migration and avoid any deterioration in mechanical properties over time, cellulose films may be coated with a hydrophobic polymer, thus preventing glycerol from leaching and water from penetrating the film. One potential candidate for coating cellulose-based films is poly(propylene carbonate) (PPC), a biodegradable aliphatic polyester synthesised from carbon dioxide and propylene oxide. A great deal of industrial and research interest has been directed towards PPC, due to the potential of its synthesis process in reducing greenhouse gas emissions and saving fossil-fuel consumption. Furthermore, its precursor propylene oxide can also potentially be synthesised from renewable sources [14]. Apart from its environmental attributes, PPC is inexpensive to produce and displays superior barrier properties and ductility. Within recent years, limited attention has been given to PPC-cellulose derivative nanocomposites, including cellulose nanocrystals [15], cellulose acetate butyrate [16] and NFC [17]. The aforementioned nanocomposites displayed enhanced mechanical properties, thermal stability and good transparency, thus demonstrating a need to further characterise, optimise and exploit these hybrid materials. Herein, the preparation and characterisation of poly(propylene carbonate)-coated NFC–CHEC–glycerol composite films is presented. A physical blending method in aqueous media was utilised to prepare composite films that possessed enhanced ductility and work-to-break, while containing as-high-as-possible concentrations of cellulosic constituents. HEC was cationised prior to blending with NFC and glycerol in order to encourage ionic complexation with the anionic cellulose nanofibrils. Given the multi-component nature of these composite films, there are several key factors that are addressed, namely: (a) The influence of NFC:CHEC ratio on material properties. (b) The influence of glycerol concentration on material properties. (c) The effectiveness of PPC coating on water resistance and mechanical properties. 2. Experimental 2.1. Materials Nanofibrillated cellulose (NFC) was provided by UPM Corporation (Helsinki, Finland), under the product name UPM Fibril Cellulose. A nominal energy consumption of 0.3 kW h kg 1 was utilised to mechanically disintegrate bleached birch kraft pulp. The birch pulp was first pre-treated with a Masuko supermasscolloider (Masuko Sangyo Ltd., Japan) to a target SR number of 90-94. Subsequently, the material was fluidized by six passes through a M7115 fluidiser (Microfluidics Corp. Newton, MA, USA) with an operating pressure of 1850 bar. The solids content of the prepared dispersion was 1.61 wt% Hydroxyethylcellulose, glycidyltrimethylammonium chloride (GTMC), poly(propylene carbonate) (PPC, average Mn  50,000), isopropanol, sodium hydroxide (NaOH), hydrochloric acid (HCl) and glycerol (anhydrous) were supplied by Sigma Aldrich, USA. 2.2. Materials preparation 2.2.1. Cationisation of HEC with GTMC HEC was cationised with GTMC using a method based on Nakamura and Sato [18]. The procedure is summarised in Scheme 1. 13.08 g of HEC was dispersed in 100 g of an 89 wt% isopropanol solution in water at 25 °C, to which 5.23 g of

330

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

Scheme 1. Cationisation of hydroxyethyl cellulose.

NaOH (20 wt%) was added and stirred for 2 h. GTAC (8.75 g, 75 wt%) was added to the mixture and stirred for 1 h at 25 °C, after which the temperature was increased to 40 °C and the mixture stirred for a further 3 h. The solution was cooled to room temperature and neutralised by HCl (4.77 g, 20 wt%), after which the cationised HEC was filtered, washed three times with acetone and dried under ambient conditions. Elemental analysis was performed using a PerkinElmer Model 2400 Series II CHNS Elemental Analyzer, while the degree of cationisation (degree of substitution by the quaternary nitrogen-containing group) was determined to be 0.46 using the method provided by Nakamura and Sato [18]. 2.2.2. Preparation of NFC–CHEC–glycerol composite films Appropriate amounts of the NFC and CHEC were dispersed in water to make 1 wt% solution and stirred at room temperature at 600 rpm for 1 h. Glycerol was added to the solution and stirred for an additional 2 h before being poured into polystyrene petri dishes and dried under ambient conditions. Film nomenclature and composition is summarised in Table 1. 2.2.3. Coating of NFC–CHEC–glycerol films with PPC NFC–CHEC–glycerol films were submersed in a 5 wt% solution of PPC in tetrahydrofuran for 15 min at room temperature. The coated films were placed in an oven at 60 °C for 15 min to remove the solvent. 2.3. Materials characterisation 2.3.1. Solid-state 13C NMR Solid state 13C cross polarisation magic angle spinning (CP MAS) NMR spectra of pure HEC and CHEC were recorded with a Bruker AVANCE-III 400 MHz spectrometer operating at 100.6 MHz for 13C. The spinning speed of samples was 8000 Hz, contact time 2 ms and delay between pulses 5 s.

331

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340 Table 1 Nomenclature and composition of NFC:CHEC:glycerol composite films. Sample name

NFC concentration (wt%)

CHEC concentration (wt%)

Glycerol concentration (wt%)

Total cellulose content (wt%)

NFC:CHEC ratio (wt%)

A5 B5 C5

71.25 47.5 23.75

23.75 47.5 71.25

5 5 5

95 95 95

75:25 50:50 25:75

A15 B15 C15

63.75 42.5 21.25

21.25 42.5 63.75

15 15 15

85 85 85

75:25 50:50 25:75

A30 B30 C30

52.5 35 17.5

175 35 52.5

30 30 30

70 70 70

75:25 50:50 25:75

2.3.2. Attenuated total reflectance infrared spectroscopy (ATR-FTIR) A Unicam Mattson 3000 FTIR spectrometer equipped with PIKE Technologies GladiATR (with diamond crystal plate) was used to characterise the chemical structure of pure HEC, CHEC and NFC–CHEC–glycerol films (both uncoated and PPC-coated). All spectra were scanned within the range 400–4000 cm 1, with a total of 16 scans and a resolution of 32 cm 1. 2.3.3. Water content Uncoated NFC–CHEC–glycerol films were weighed, placed in an oven at 120 °C for 3 h and subsequently re-weighed. All specimens were characterised in duplicate. 2.3.4. Scanning electron microscopy (SEM) A Zeiss SIGMA VP scanning electron microscope operating at 300 V was used to characterise the morphology of both uncoated and PPC-coated NFC–CHEC–glycerol films. Samples were mounted to the specimen holder using carbon tape. 2.3.5. Contact angle Water contact angles of PPC-coated and uncoated NFC–CHEC–glycerol composites were determined from the surface of the films. A Tantec AS Contact Angle Metre was used to measure the static contact angles of NFC–CHEC
glycerol films (both PPC-coated and uncoated) at room temperature. Values were measured within 10 s of releasing the water drop (100 lL), with results presented being the average of three measurements. 2.3.6. Stress–strain Stress–strain analysis was performed using an Instron Universal Testing Instrument, Model 33R4204 with a 100 N static load attached. A strain rate of 0.5 mm min 1 was applied to each sample (average dimensions: 10.00  5.30  0.05 mm3) at 23 °C and 50% relative humidity. Wet strength characterisation of PPC-coated NFC–CHEC–glycerol films was performed using specimens immersed in distilled water for 24 h, with the dimensions being measured from dry specimens. Results presented are the average of five measurements. 3. Results and discussion 3.1. Structure and composition of cationic-HEC The FTIR spectra of pure and cationised HEC are presented in Fig. 1. Pure HEC displayed several peaks characteristic of its structure; hydroxyl (AOH) stretching at 3378 cm 1, stretching of alkyl (ACH2 and ACH) groups at 2882 cm 1, ether linkage (ACAOACA) stretching at 1038 cm 1 and AOH bending at 1356 cm 1 [19]. The peak at 890 cm 1 is derived from b-glucosidic linkages between the glucose units [20] while the small band at 1645 cm 1 can be attributed to the bending of naturally adsorbed water. CHEC retained these characteristic peaks, however a prominent, sharp peak emerged at 1447 cm 1. This was attributed to the stretching vibration of CAN on the quaternary ammonium ion [21]. The peak corresponding to adsorbed water at 1645 cm 1 has almost completely diminished, appearing as a faint shoulder of the main peak at 1447 cm 1, while the AOH peak at 3378 cm 1 experienced a similar reduction in intensity. This indicates there are less vacant hydroxyl groups able to interact with water molecules, most likely due to etherification of hydroxyl groups that occurred during the cationisation reaction. To further confirm the cationisation of HEC, 13C NMR spectra of pure and modified HEC were characterised and are presented in Fig. 2. Pure HEC exhibited characteristic peaks associated with its D-glucose backbone. The cationised-HEC likewise displayed a spectrum detailing the pure HEC structure, however a notable difference was the emergence of a peak at 57.1 ppm. This was attributed to carbon atoms of the N-trimethylated group [22–24]. In order to determine whether the etherification reaction between HEC and GTAC was successful, the peak corresponding to substituted C6s can provide information. Zhou et al. [25] assigned a small peak at approximately 70 ppm to that of substituted C6 (ie. where the ‘R’ group is

332

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

Fig. 1. FTIR spectra of pure HEC and CHEC.

Fig. 2. Solid state

13

C NMR spectra; (i) pure HEC and (ii) cationic HEC.

not a hydrogen atom nor CH2CH2OH). This peak does not appear within pure HEC since only CH2CH2OH groups are present and they are displayed at 72.5 ppm. Within the CHEC spectrum, a shoulder develops at 70.6 ppm. It is unlikely that the presence of unreacted GTAC would enhance the C6s peak of HEC. Thus, it implies successful ether-linkage formation between the ethoxy group of HEC and epoxide ring of GTAC. 3.2. Interaction and bonding between NFC and CHEC IR spectroscopy has been utilised to confirm the occurrence of electrostatic interactions and ionic interactions [26–28], since stronger bonds display a higher force constant and vibrate at higher frequencies (wavenumbers). Thus ATR-FTIR was also used to evaluate the nature of bonding and interactions within the composite films. The FTIR spectra of uncoated NFC–CHEC–glycerol films containing 30 wt% glycerol content are presented in Fig. 3. Similar composites were prepared with pure, non-cationised HEC to determine the effect of cationic species within the films. When comparing the spectra of composite A30 containing non-cationic- and cationic HEC, a shift to higher wavenumber is observed for the CAN stretching vibration of the quaternary ammonium ion (1447 cm 1) and for hydroxyl stretching (3378 cm 1). Similar observations are seen for B30 composite spectra. Since the only difference between these blends was the presence of pure or cationic HEC, the shift to higher wavelengths can be attributed to stronger bond linkages deriving from ionic interactions between the respective ionic groups. When the C30 composite (that is, the composite with the highest HEC concentration) spectra are compared, only the hydroxyl peaks shifted to higher wavenumbers, while the quaternary ammonium ion peaks remained stationary. At elevated CHEC concentrations, the increased number of quaternary ammonium ions results in a greater degree of electrostatic repulsion between the cationic groups, leading to swelling of CHEC chains [29]. This increases polymer solubility [30], the

333

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

Fig. 3. FTIR spectra of various NFC:CHEC:glycerol composites with 30 wt% glycerol. Solid lines correspond to films containing cationic HEC while dashed lines represent films containing pure HEC.

Table 2 Water content, absorption and swelling properties of NFC:CHEC:glycerol films. Sample

Water content (%)

Static hwater – uncoated films (°)

Static hwater – coated films (°)

A5 B5 C5

12 ± 4 22 ± 3 29 ± 6

58 ± 3 62 ± 2 58 ± 4

70 ± 5 73 ± 4 68 ± 4

A15 B15 C15

21 ± 2 31 ± 4 39 ± 4

49 ± 2 56 ± 3 41 ± 3

78 ± 3 77 ± 5 61 ± 4

A30 B30 C30

64 ± 3 67 ± 3 53 ± 5

32 ± 4 43 ± 3 28 ± 4

75 ± 4 85 ± 8 66 ± 5

hydrophilicity of cellulosic materials and the amount of water able to interact with the CHEC, NFC and glycerol, reducing the likelihood of quaternary ammonium–hydroxyl complexation. To further probe the influence of adsorbed water, the water content of the films is presented in Table 2. The water–adsorption ability of glycerol at elevated humidities is well-established [31,32], thus the water content of the films increased with glycerol concentration. At a constant glycerol loading of 5 wt%, increasing the CHEC component increased the water content from 12% (specimen A5) to 29% (specimen C5). Similar behaviour was observed for other sets of films, with composite B30 displaying a maximum water content of 67%. Therefore, the proposed mechanism of repulsion-induced swelling and subsequent water adsorption appears to be plausible. Due to the multi-component nature of the NFC–CHEC–glycerol composites, it is apparent that various forms of interaction are occurring between the components. The data indicates that at CHEC concentrations up-to-and-including 47.5 wt%, electrostatic interactions between ions are present within the films. At higher CHEC loadings, the occurrence of ionic complexation is significantly diminished, with hydrogen bonding being the dominant mode of interaction. Regarding interactions between NFC and glycerol, other NFC segments and/or CHEC hydroxyl/ethoxyl groups, hydrogen bonding is also believed to be the dominant mode of interaction. 3.3. Film morphology and coating efficiency 3.3.1. Scanning electron microscopy Scanning electron micrographs of selected NFC–CHEC–glycerol films are presented in Fig. 4. Within all compositions, the fibrilar structure of NFC is evident. As the glycerol concentration is increased the fibrilar network structure becomes less evident, in particular at maximum glycerol concentrations of 30 wt% (Fig. 4b). It is apparent that the glycerol effectively coats NFC and penetrates into CHEC domains within the films, indicating sufficient interaction between the phases. Within films containing 5 and 15 wt% glycerol (Fig. 4a), round-shaped CHEC aggregates are noticeable. Similar phenomena have been observed within bacterial cellulose- and NFC films [9] compounded with pure HEC. The proposed mechanism of interaction implies that within composites where NFC is the larger component, HEC coats the nanocellulose fibrils while excess HEC may aggregate. Within films consisting of higher concentrations of CHEC (Fig. 4c), NFC clusters were embedded throughout the

334

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

Fig. 4. Scanning electron micrographs of NFC–CHEC–glycerol films; (a) A15, (b) A30, (c) B15, and (d) A15 coated with PPC.

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

335

soft CHEC matrix component. As anticipated, the prominence of CHEC agglomeration is visibly higher at increased loadings, as indicated by the more numerous white aggregates. The SEM micrograph of the A15 NFC–CHEC–glycerol composite coated with PPC is presented in Fig. 4d. In contrast to their non-coated counterparts, all hybrid films coated with PPC exhibited a smooth morphology, with no noticeable cellulose nanofibrils or hydroxyethylcellulose agglomerates. However, given that the non-coated NFC–CHEC–glycerol films possessed small voids throughout their morphology (refer Fig. 4), it is possible that small fractions of PPC may have been adsorbed within the NFC–CHEC–glycerol films. Similarly, following PPC-coating, all films displayed small cracks or voids throughout the surface morphology. Both nanofibrillated cellulose [33] and poly(propylene carbonate) [34] can experience shrinkage due to solvent evaporation, with phase separation also being able to contribute. Thus, despite the overall smooth surface coating, PPC-coated films contained nano-dimension voids throughout their morphology – leading to potential implications on mechanical and water barrier properties. The cross-section micrographs of selected composites coated with PPC are presented in Fig. 5. At increased NFC concentrations (NFC:CHEC ratio 75:25, Fig. 5a) a smooth coating-layer of PPC is visible with a thickness of 4.5–6 lm. The PPC readily coated the cellulose nanofibrils at the interface of the composite and coating, suggesting sufficient adhesion via hydrogen bonding was achieved. In contrast, composites containing a higher hydroxyethylcellulose concentration (NFC:CHEC ratio 25:75, Fig. 5b) displayed a crimpled PPC coating layer with no visible wetting at the interface. This indicates that although a degree of adhesion and interaction was present between PPC and CHEC, the poly(propylene carbonate) coating adheres and wets NFC more effectively.

3.3.2. ATR-FTIR ATR-FTIR was utilised to characterise the surface chemistry of the PPC-coated nanocomposite films, the spectra of selected composites shown in Fig. 6. All films spectra were similar in appearance, displaying characteristics peaks corresponding to PPC [35–38]; 1743 cm 1 (C@O stretching), 1235 cm 1 (OACAO stretching), 2972 cm 1 (CH2 stretching), 1160 cm 1 (ACAOA stretching within the ACHAOA group) and 1067 cm 1 (ACAOA stretching within the AOAC@O group). A low, broad peak is visible at 3400 cm 1, corresponding to the stretching vibrations of AOH within the NFC–CHEC–glycerol films. However, given the relatively low peak intensity in relation to other bands, the FTIR data strongly suggests that the bulk of the film surface is PPC. This is in good correlation with the SEM images, indicating successful coating of the nanocomposite films with PPC.

Fig. 5. Cross-section micrographs of PPC-coated NFC–CHEC–glycerol films; (a) A15 and (b) C15.

336

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

Fig. 6. FTIR spectra of selected PPC-coated NFC–CHEC–glycerol films.

3.4. Wetting and barrier properties of PPC-coated NFC–CHEC glycerol films The contact angle (h) properties of the uncoated and PPC-coated NFC–CHEC–glycerol films are summarised in Table 2. Uncoated film A5 displayed a h value of 43°, similar to the values obtained for pure NFC (43°) and pure HEC (45°) films. As the glycerol concentration was increased to 15 wt% (A15), h dropped to 32°. Further addition of glycerol (up to 30 wt%) reduced the contact angle to 28°. This behaviour is indicative of glycerol’s high affinity for water and was anticipated, given glycerol’s tendency to migrate towards the surface in cellulosic films [39]. Increasing the NFC:CHEC ratio up to 50:50 within the uncoated films increased h values. Further addition of CHEC (NFC:CHEC ratio of 25:75) caused the h value to decrease, yielding similar values to composites containing a NFC:CHEC ratio of 75:25. This behaviour was observed for the uncoated films, regardless of glycerol concentration. The structure of non-modified HEC possess three ‘R’ substituents attached to the pyranose ring, which can be either hydrogen atoms or an ethoxy functional group. In the case of CHEC, one or more of these ‘R’ groups may be the bonded GTAC unit. Both pure and modified HEC have less functional groups able to interact and adsorb water than on NFC (which, in contrast, has three hydroxyl groups attached to its backbone ring). This leads to the initial increase in contact angle values. The reduction in hydrophobicity observed at NFC:CHEC ratios of 25:75 may be attributed to the increased presence of quaternary ammonium groups that may interact strongly with both glycerol [40] and water molecules [41]. However, a simpler explanation may be due to the CHEC being the primary matrix component in the composites and both NFC and HEC displaying similar h values. The PPC-coated films displayed higher h values than their non-coated equivalents, with specimen B30 displaying a maximum of 85°. Effects from glycerol and CHEC concentration within the PPC-coated films could not be clearly distinguished, partly due to effective coating of the composite film surface as confirmed by SEM (refer Fig. 4d). Any discrepancies within the contact angle values most probably originate from the presence of cracks and voids within the PPC-coated films and subsequent film-surface roughness. The increased h values correlate well with the ATR-FTIR and SEM data, confirming the successful coating of the films with PPC and demonstrating its limited ability in reducing water wetting on cellulose-based films. Films (both PPC-coated and non-coated) were kept immersed in water to determine the long-term water stability behaviour. Non-coated composite films began to disintegrate almost instantly following their immersion in water. This was expected due to the hydrophilic nature of the composite components. PPC-coated films retained their form and structural integrity following 30 min of immersion, being able to be manipulated and handled. However, the films were softer and more delicate when compared to before exposure to water. After 45 min of immersion the films began to soften further, while following 1 h the films could easily be broken apart with gentle force. Furthermore, the PPC-coating layer separated from the NFC–CHEC–glycerol composite substrate after 1 h of submersion in water. The adhesion mechanism between the NFC–CHEC–glycerol nanocomposite and the PPC coating is most-probably hydrogen bonding between the carbonyl group of PPC and the various hydroxyl groups within the composite. However, defects or voids within specimens, as well as areas with little or even no PPC coating may allow water to enter the films. Prolonged exposure to water can disrupt interactions between the PPC and NFC–CHEC–glycerol constituents, as preferential hydrogen bonding with water molecules occurs. Water-induced swelling is also highly probable at elevated immersion times. The result is peeling of the PPC coating and rapid deterioration in NFC–CHEC–glycerol structure. Although the PPC-coated films effectively repealed water at short exposure times and low volumes, physical interaction-base coating-substrate behaviour is not suitable for long-term stability or applications where long exposure times or high concentrations of water are required.

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

337

3.5. Tensile mechanical behaviour 3.5.1. Uncoated NFC–CHEC–glycerol films To determine the influence of ionic interactions on mechanical behaviour, uncoated composite films were prepared using both non-modified- and cationised-HEC. All other concentrations and processing conditions were kept identical. The stress– strain curves of uncoated composites containing non-modified- and cationised-HEC (glycerol concentration 5 wt%) are presented in Fig. 7 while the tensile data is summarised in Table S1 of the Supplementary Data. Distinct differences in tensile behaviour are evident for pure- and cationised-HEC composites, with the later displaying larger tensile moduli and strength values, while also exhibiting greater brittleness than their pure-HEC blended counterparts. Similar behaviour was described by Feng et al. [42] with polyelectrolyte complex films consisting of polyvinyl amine and carboxymethyl cellulose. Increased strength and brittleness observed in the dry-state was attributed to strong intermolecular bonds (ionic bonding) from polar groups that would otherwise hydrogen bond or interact with water. This behaviour, in conjunction with FTIR and water absorption data, confirms that ionic interactions between the quaternary ammonium cation of HEC and hydroxyls of NFC are occurring. However, it is probable that the main mode of interaction stems from hydrogen bonding between nanofibrils. Increasing the CHEC concentration had a drastic influence on film ductility. Within composites containing 5 wt% glycerol, the strain at break increased from 14% (NFC:CHEC ratio 75:25) to 86% (NFC:CHEC 25:75). The values are remarkable considering the concentration set was composed of 95 wt% cellulose material. Within NFC films, the ductility and elongation behaviour is significantly influenced by interfibril friction and the ease at which these nanofibrils can experience slippage [9,43]. CHEC acts as a lubricant for NFC, coating the nanofibrils and allowing them to undergo interfibril slippage and possible alignment in the direction of deformation while also limiting interactions between cellulose nanofibrils. As anticipated, the moduli and maximum strength values decreased with increased loading of soft-segments of CHEC. Furthermore, CHEC agglomerates observed in SEM micrographs may act as stress-concentrators and exhibit a negative effect on modulus and strength values. At glycerol concentrations of 15 and 30 wt% (Fig. 7b and c, respectively), films composed of NFC:CHEC ratios of 25:75 displayed signs of strain hardening. Within polymer films, this behaviour is associated with chain orientation and alignment under load [44]. The ability of CHEC to coat NFC facilitates fibril slippage, straightening and subsequent alignment/reorientation during deformation [9]. Similarly, increasing the glycerol concentration had an evident impact on film ductility. At a NFC:CHEC ratio of 25:75 and 30 wt% glycerol concentration (Fig. 7c), a maximum strain-at-break value of 141% was obtained. The levels are amongst the highest obtained for cellulose-based films and impressive considering the film was comprised of 70 wt% materials derived from cellulose. Plasticisers such as glycerol have readily been utilised to enhance elasticity and ductility within cellulose and other biopolymer-based films. As with HEC, glycerol molecules reduced friction between NFC fibrils and limit fibril–fibril interactions. However, the aforementioned phenomena and reduced crystallinity that are associated with glycerol incorporation prove detrimental to other mechanical properties, namely material strength and moduli. As anticipated, higher loadings of glycerol resulted in films with diminished moduli and strength values. In addition to film ductility, the tear strength properties were the second major mechanical property of importance. Tear strength is related to the toughness or the energy required to fracture a material. Thus, the work-to-break was obtained from the area underneath the stress–strain curves of the films, to provide an approximate tear strength value. The values are summarised in Table S1 of the Supplementary Information. At a glycerol concentration of 5 wt%, the NFC:CHEC 75:25 composite displayed a work-to-break value of 4.9 MJ m3. Increasing the CHEC concentration to 50% (relative to NFC) increased values to 10.8 MJ m3 while a maximum value of 12.3 MJ m3 was obtained at CHEC loadings of 75%. Likewise, at glycerol concentrations of 15 and 30 wt%, the work-to-break increased with CHEC concentration, reaching a maximum value of 16.9 MJ m3 for specimen C15. Increasing the glycerol concentration from 5 to 15 wt% had a positive effect on work-to-break, while higher loadings of 30 wt% yielded films with the lowest work-to-break values. The primary factor influencing tear behaviour is the slippage of NFC nanofibrils which results in orientation and alignment. As the general deformation mechanism within the NFC:CHEC:gly composite films is based around the aforementioned phenomena, plastic deformation of the actual nanofibrils is prevented during load application. Thus an increase in the energy required to tear the films accompanies the enhanced ductility and stretchability. A comparison video demonstrating the difference in tear behaviour of pure NFC and specimen B15 is presented in the Supplementary Material. An initial balance between strength and ductility can be achieved at glycerol concentrations up to and including 15 wt%, with maximum strength ranges of 25–45 MPa and strain at break values of 16–109% being readily attainable. Although higher tear strength values for NFC-HEC composite systems have been reported in literature [8,9], the NFC–CHEC–gly composites do displayed improved work-to-break values than pure NFC and HEC and are comparable with other NFC-based systems. Thus, NFC:CHEC:gly composites may find use in a variety of applications, since brittleness and poor-handleability of nanocellulose films has been one critical factor preventing their large scale utilisation. Higher loadings of glycerol (30 wt%) further enhanced fibril slippage, but severely diminished any reinforcement available from fibril–fibril interactions. While the strain at break values were high, it is at the expense of moduli, strength and toughness. By preparing a series of compositions from one extreme concentration to another, optimisation and fine-tuning of component proportions can be utilised. Although multiple methods can be utilised to prepare composite films, one benefit of using solution casting is that species dissolved in water (including glycerol, poly(vinyl alcohol), etc.) can be retained in the composite structure. However, the opportunity for utilising other compounds or materials may provide retention of strength and toughness while simultaneously accommodating enhanced ductility and elongation.

338

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

Fig. 7. Stress–strain curves of uncoated NFC–CHEC–glycerol films; (a) 75:25:5 (solid line specimens contain CHEC while dotted line specimens contain pure HEC), (b) 75:25:15, and (c) 75:25:30.

3.5.2. PPC-coated NFC–CHEC–glycerol films The stress–strain data of the PPC-coated NFC–CHEC–glycerol films is summarised in Table S2 of the Supplementary Data. Composite films coated with PPC displayed superior moduli and strength values than their non-coated equivalents. A5 exhibited a maximum modulus of 2270 MPa and strength of 81.4 MPa. However, the influence of PPC coating on moduli values was more evident within composites containing a higher NFC fraction. As the CHEC concentration was increased, the difference in moduli values between coated and non-coated composites became smaller. This phenomenon may be attributed to the intrinsic mechanical properties of PPC and CHEC. Within NFC–CHEC–glycerol composites where NFC is the larger cellulosic component, PPC may have a greater impact on mechanical behaviour due to possessing a higher stiffness than CHEC. Furthermore, microscopy confirmed effective wetting and adhesion between NFC and PPC (refer Fig. 5a). However, for composites with an increased CHEC component, the primary reinforcement derives from NFC. Given the extremely tough

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340

339

and strong nature of NFC, the influence and impact of PPC within these composites is considerably diminished by comparison. The influence of glycerol and CHEC concentration within the PPC-coated films displayed identical trends as their non-coated equivalents (ie, reduced strength and moduli with increasing concentration). This indicates that although PPC provided some additional reinforcement, the composition of the composites and influence of the various components (NFC, CHEC and glycerol) primarily determine tensile behaviour. Generally, PPC-coated NFC–CHEC–glycerol composites displayed inferior strain at break and work-to-break values than their uncoated counterparts. This was especially evident at increased CHEC concentrations (NFC:CHEC ratios of 50:50 and 25:75), where PPC-coated films exhibited lower strain at break and strength values than their uncoated counterparts. As mentioned previously, the enhanced ductility within the composites is attributed to cellulosic fibril slippage, straightening and subsequent alignment/reorientation during deformation as a result of coating with CHEC and glycerol [9]. The various hydrogen bonds and dipole-ion interactions formed between NFC, CHEC and PPC can lead to the creation of ‘anchors’ or points which reduce fibril mobility during deformation. While this does lead to modulus and strength enhancement, this is at the expense of ductility and tear strength. Furthermore, the presence of voids within composite films following PPC coating (refer Fig. 4d) may increase the deterioration of mechanical properties, in particular when a soft matrix (such as CHEC) is the dominant component. In this case, the NFC fibrils dispersed throughout the CHEC provide more effective reinforcement than PPC. This highlights the limitations of PPC-coating, which can be useful for certain applications rather than a general approach. Despite the reduction, the coated composites continued to show impressive strain at break values (for NFC standards). 4. Conclusion Nanofibrillated cellulose (NFC) was successfully blended with cationic hydroxyethylcellulose (CHEC) and glycerol to yield hybrid nanocomposite films. Cationisation of HEC was confirmed via NMR and FTIR, while amounts of non-cationised HEC were also present. Composite films possessed a fibrillar network structure, with the NFC readily coated with segments of non-functionalised HEC and glycerol. Ionic interactions were also evident between cationic HEC and nanofibrillated cellulose, the enhanced bonding contributing towards moduli, strength and work-to-break values. Increasing glycerol and CHEC concentration enhanced film ductility, due to the lubricating effect of the components on the cellulose nanofibrils. However, as is expected in composites, the enhanced strain at break values were at the expense of strength and toughness. Poly(propylene carbonate) readily coated the surface of the NFC–CHEC–glycerol films, as confirmed via SEM. Despite some enhancement in water stability and hydrophobicity for an acute time period, longer immersion times caused the films to disintegrate. This was attributed to voids formed in the films during solvent evaporation and the weak hydrogen-bond holding the PPC coating to the NFC–CHEC–glycerol film surface. PPC provided some enhancement in moduli and strength values at lower CHEC concentrations; however the mechanical behaviour within the composites was primarily dictated by the primary components (nanofibrillated cellulose, cationic hydroxyethyl cellulose and glycerol). The use of hydroxyethylcellulose and glycerol in conjunction with NFC demonstrated the ability to prepare extremely ductile cellulose-based and cellulose-majority films utilising simple casting techniques. However, the issues of reduced strength, water stability and plasticiser leaching remain. Similarly, although PPC showed promise as a coating material for these composites, stronger bonding and interaction mechanisms ought to be implemented to further enhance the properties of these hybrid films. Acknowledgments The authors wish to thank Tiia Juhala for performing elemental analysis. This work made use of Aalto University Bioeconomy and Aalto University Nanomicroscopy Center (Aalto-NMC) facilities. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.eurpolymj.2015.06.019. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]

D.M. Fox, J. Lee, M. Zammarano, D. Katsoulis, D.V. Eldred, L.M. Haverhals, P.C. Trulove, H.C. De Long, J.W. Gilman, Carbohydr. Polym. 88 (2012) 847–858. A.N. Nakagaito, H. Yano, Appl. Phys. A 80 (2005) 155–159. E. Trovatti, S.C.M. Fernandes, L. Rubatat, D.d.S. Perez, C.S.R. Freire, A.J.D. Silvestre, C.P. Neto, Compos. Sci. Technol. 72 (2012) 1556–1561. C. Aulin, G. Salazar-Alvarez, T. Lindstrom, Nanoscale (2012) 6622–6628. K. Larsson, L.A. Berglund, M. Ankerfors, T. Lindström, J. Appl. Polym. Sci. 125 (2012) 2460–2466. N. Hansen, T. Blomfeldt, M. Hedenqvist, D. Plackett, Cellulose 19 (2012) 2015–2031. M. Österberg, J. Vartiainen, J. Lucenius, U. Hippi, J. Seppälä, R. Serimaa, J. Laine, ACS Appl. Mater. Interfaces 5 (2013) 4640–4647. H. Sehaqui, Q. Zhou, L.A. Berglund, Soft Matter 7 (2011) 7342–7350. H. Sehaqui, S. Morimune, T. Nishino, L. Berglund, Biomacromolecules 13 (2012) 3661–3667. Q. Zhou, E. Malm, H. Nilsson, P.T. Larsson, T. Iversen, L.A. Berglund, V. Bulone, Soft Matter 5 (2009) 4124–4130. K. Luo, J. Yin, O.V. Khutoryanskaya, V.V. Khutoryanskiy, Macromol. Biosci. 8 (2008) 184–192.

340 [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44]

S. Spoljaric et al. / European Polymer Journal 69 (2015) 328–340 R. Yang, R. Shi, S. Peng, D. Zhou, H. Liu, Y. Wang, Electrophoresis 29 (2008) 1460–1466. K. Wang, L. Ye, J. Macromol. Sci., Part B 53 (2014) 149–161. Q. Chen, E.J. Beckman, Green Chem. (2008) 934–938. X. Hu, C. Xu, J. Gao, G. Yang, C. Geng, F. Chen, Q. Fu, Compos. Sci. Technol. 78 (2013) 63–68. C. Xing, H. Wang, Q. Hu, F. Xu, X. Cao, J. You, Y. Li, Carbohydr. Polym. 92 (2013) 1921–1927. D. Wang, J. Yu, J. Zhang, J. He, J. Zhang, Compos. Sci. Technol. 85 (2013) 83–89. H. Nakamura, S. Sato, US Patent 6197950 B1, 2001. A. El idrissi, S. El barkany, H. Amhamdi, A. Maaroufi, J. Appl. Polym. Sci. 127 (2013) 3633–3644. F.W. Langkilde, A. Svantesson, J. Pharm. Biomed. Anal. 13 (1995) 409–414. K. Wang, L. Ye, Polym.-Plast. Technol. Eng. 49 (2010) 807–811. S. de Matos Fernandes, PhD Thesis, University of Averio, 2010, p. 227. D. de Britto, O.B.G. de Assis, Int. J. Biol. Macromol. 41 (2007) 198–203. R. Komban, R. Beckmann, S. Rode, S. Ichilmann, A. Kühnle, U. Beginn, M. Haase, Langmuir 27 (2011) 10174–10183. Q. Zhou, L. Zhang, M. Li, X. Wu, G. Cheng, Polym. Bull. 53 (2005) 243–248. H. Binder, G. Lindblom, Biophys. J. 87 (2004) 332–343. J.F. Luna-Martínez, D.B. Hernández-Uresti, M.E. Reyes-Melo, C.A. Guerrero-Salazar, V.A. González-González, S. Sepúlveda-Guzmán, Carbohydr. Polym. 84 (2011) 566–570. H. Staroszczyk, K. Sztuka, J. Wolska, A. Wojtasz-Paja˛k, I. Kołodziejska, Spectrochim. Acta Part A: Mol. Biomol. Spectrosc. 117 (2014) 707–712. A.V. Svensson, L. Huang, E.S. Johnson, T. Nylander, L. Piculell, ACS Appl. Mater. Interfaces 1 (2009) 2431–2442. H. Lee, M.Q. Yee, Y.Y. Eckmann, N.J. Hickok, D.M. Eckmann, R.J. Composto, J. Mater. Chem. 22 (2012) 19605–19616. A. Dufresne, D. Dupeyre, M.R. Vignon, J. Appl. Polym. Sci. 76 (2000) 2080–2092. P. Rachtanapun, W. Tongdeesoontorn, Asian J. Food Agro-Ind. 2 (2009) 478–488. M. Manninen, I. Kajanto, J. Happonen, J. Paltakari, Nord. Pulp Pap Res. J. 26 (2011) 297–305. L.F. Francis, A.V. Mccormick, D.M. Vaessen, J.A. Payne, J. Mater. Sci. 37 (2002) 4717–4731. X. Jing, M.R. Salick, T. Cordie, H. Mi, X. Peng, L. Turng, Ind. Eng. Chem. Res. 53 (2014) 9391–9400. X. Ma, P.R. Chang, J. Yu, N. Wang, Carbohydr. Polym. 71 (2008) 229–234. G.A. Luinstra, E. Borchardt, in: B. Rieger, A. Kunkel, G.W. Coates, R. Reichardt, E. Dinjus, T.A. Zevaco (Eds.), Material Properties of Poly(Propylene Carbonates), Springer, Berlin Heidelberg, 2012, pp. 29–48. J. Gao, B. Hongwei, Xin Zhou, Guanghui Yang, Chenlong Xu, Qin Zhang, Feng Chen, Qiang Fu, Nanotechnology 25 (2014) 025702. C. Mu, J. Guo, X. Li, W. Lin, D. Li, Food Hydrocolloids 27 (2012) 22–29. A.P. Abbott, P.M. Cullis, M.J. Gibson, R.C. Harris, E. Raven, Green Chem. 9 (2007) 868–872. H. Chang, J. Jiang, T. Chen, H. Wang, L.Y. Li, W. Hung, S.H. Lin, Phys. Chem. Chem. Phys. 15 (2013) 12734–12741. X. Feng, R. Pelton, M. Leduc, Ind. Eng. Chem. Res. 45 (2006) 6665–6671. M. Wang, A. Olszewska, A. Walther, J. Malho, F.H. Schacher, J. Ruokolainen, M. Ankerfors, J. Laine, L.A. Berglund, M. Österberg, O. Ikkala, Biomacromolecules 12 (2011) 2074–2081. M. Wendlandt, T.A. Tervoort, U.W. Suter, J. Polym. Sci., Part B: Polym. Phys. 48 (2010) 1464–1472.