Article pubs.acs.org/Biomac
Xyloglucan-Functional Latex Particles via RAFT-Mediated Emulsion Polymerization for the Biomimetic Modification of Cellulose Fiona L. Hatton,† Marcus Ruda,‡ Muriel Lansalot,§ Franck D’Agosto,§ Eva Malmström,† and Anna Carlmark*,† †
KTH Royal Institute of Technology, School of Chemical Science and Engineering, Department of Fibre and Polymer Technology, Teknikringen 56, SE-100 44 Stockholm, Sweden ‡ CelluTech AB, Teknikringen 38, SE-114 28 Stockholm, Sweden § Université de Lyon, Univ Lyon 1, CPE Lyon, CNRS, UMR 5265, C2P2 (Chemistry, Catalysis, Polymers and Processes), Team LCPP, Bat 308F, 43 Bd du 11 Novembre 1918, 69616 Villeurbanne, France S Supporting Information *
ABSTRACT: Herein, we report a novel class of latex particles composed of a hemicellulose, xyloglucan (XG), and poly(methyl methacrylate) (PMMA), specially designed to enable a biomimetic modification of cellulose. The formation of the latex particles was achieved utilizing reversible addition− fragmentation chain transfer (RAFT) mediated surfactant-free emulsion polymerization employing XG as a hydrophilic macromolecular RAFT agent (macroRAFT). In an initial step, XG was functionalized at the reducing chain end to bear a dithioester. This XG macroRAFT was subsequently utilized in water and chain extended with methyl methacrylate (MMA) as hydrophobic monomer, inspired by a polymerization-induced self-assembly (PISA) process. This yielded latex nanoparticles with a hydrophobic PMMA core stabilized by the hydrophilic XG chains at the corona. The molar mass of PMMA targeted was varied, resulting in a series of stable latex particles with hydrophobic PMMA content between 22 and 68 wt % of the total solids content (5−10%). The XG-PMMA nanoparticles were subsequently adsorbed to a neutral cellulose substrate (filter paper), and the modified surfaces were analyzed by FT-IR and SEM analyses. The adsorption of the latex particles was also investigated by quartz crystal microbalance with dissipation monitoring (QCM-D), where the nanoparticles were adsorbed to negatively charged model cellulose surfaces. The surfaces were analyzed by atomic force microscopy (AFM) and contact angle (CA) measurements. QCMD experiments showed that more mass was adsorbed to the surfaces with increasing molar mass of the PMMA present. AFM of the surfaces after adsorption showed discrete particles, which were no longer present after annealing (160 °C, 1 h) and the roughness (Rq) of the surfaces had also decreased by at least half. Interestingly, after annealing, the surfaces did not all become more hydrophobic, as monitored by CA measurements, indicating that the surface roughness was an important factor to consider when evaluating the surface properties following particle adsorption. This novel class of latex nanoparticles provides an excellent platform for cellulose modification via physical adsorption. The utilization of XG as the anchoring molecule to cellulose provides a versatile methodology, as it does not rely on electrostatic interactions for the physical adsorption, enabling a wide range of cellulose substrates to be modified, including neutral sources such as cotton and bacterial nanocellulose, leading to new and advanced materials.
■
functionalities.2 Modification of cellulose with polymers can be achieved by covalent attachment via grafting-from or grafting-to methods5−7 or by physical adsorption exploiting electrostatic interactions, hydrogen bonding, or van der Waals forces.8 One of the largest industries based around cellulose, the pulp and paper industry, has utilized polyelectrolytes for decades to modify cellulose fibers via physical adsorption. This physical adsorption, or physisorption, is achieved through electrostatic
INTRODUCTION
The incorporation of cellulose in biocomposites has been a research area of significant interest in recent years due to the natural abundance, physical properties and biorenewability of cellulose, the main constituent of plant cell walls.1 The growing demand for biobased materials to replace fossil fuel-derived materials has most likely stimulated this interest into biocomposite research.2−4 However, due to the hydrophilic nature of cellulose, it has poor compatibility with hydrophobic matrix polymers commonly used in composites. Therefore, the modification of cellulose is necessary to impart desired properties, such as improving miscibility with hydrophobic polymers, barrier properties and incorporating specific © XXXX American Chemical Society
Received: January 11, 2016 Revised: February 24, 2016
A
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules interactions between a cationic polyelectrolyte and anionic cellulose substrates.9 The introduction of negative charges to cellulose is accomplished through oxidation of the primary hydroxy groups present on each repeating unit (anhydro-Dglucose) of the cellulose chain, for example, by carboxylation or sulfonation, therefore, enabling electrostatic interactions to occur with cationic species.1 Previous work in our group has focused on the modification of cellulose with polymers both through covalent10−13 and electrostatic14−17 interactions. Targeting physical adsorption through electrostatic interactions, tailor-made block copolymers have been utilized, which contain a quaternized poly(dimethylaminoethyl methacrylate) (PDMAEMA) block14−16 or a poly((acrylamidopropyl) trimethylammonium chloride) (PAPTAC) block.17 Cationic latexes are well-known for the modification of cellulose. Studies into the reinforcement of paper have been reported utilizing cationic latexes of poly(styrene-co-butadiene).18−20 Cationic latexes composed of poly(styrene-co-nbutyl acrylate) were used to prepare nanocomposite films with cellulose whiskers to improve mechanical properties.21 The same type of latexes have been studied for their adhesive properties to cellulose22 and their adsorption to cotton fibers.23 Core−shell latex particles of varying compositions of styrene, nbutyl acrylate, methyl methacrylate, and 2-ethylhexyl acrylate were also investigated in their adsorption to cellulose fibers, increasing the hydrophobicity and enabling incorporation into a polypropylene-based biocomposite.24−26 The general principle is always to design latex particles carrying cationic charges at the surface that can interact with the negatively charged cellulose and a hydrophobic core polymer of interest for the targeted cellulose modification. First described by Hawkett and co-workers in 2005,27 polymerization-induced self-assembly (PISA) describes the chain extension, in water, of hydrophilic polymer chains, synthesized by controlled radical polymerization, with a hydrophobic monomer.28−31 When the growth of the hydrophobic segment is well-controlled, the resulting latex nanoparticles are exclusively composed of amphiphilic block copolymers. The PISA process has been well described previously, including the one-pot synthesis of latex particles in water.32−34 While not only an excellent tool to design amphiphilic block copolymers in high yield directly in water, PISA is also a very efficient method to design the particle surface at will. For example, a variety of original, carbohydrate functional nanoparticles have already been obtained,35−37 including alginate-decorated nanoparticles.38 With the aim of modifying the surface of cellulose, we have recently reported the synthesis of poly(methyl methacrylate) (PMMA) latex particles, utilizing the PISA mechanism, stabilized by P(DMAEMA-co-methacrylic acid (MAA)) macroRAFT segments.39 The manipulation of pH led to the protonation of the DMAEMA repeating units, resulting in cationic latexes which were subsequently investigated for their adsorption to a negatively charged cellulose substrate as followed by quartz crystal microbalance with dissipation monitoring (QCM-D). As mentioned previously, the use of cationic polyelectrolytes and latexes to modify cellulose has been well studied,9,40 including adsorption to cellulosic substrates such as nanocelluloses.14,41,42 Although the use of nonpolyelectrolytes, or neutral polymers, to modify cellulose has not been as well investigated, Perrier and co-workers have studied the adsorption of synthetic polymers with varying architectures to cellulose, including glucose units for cellulose recognition.43
When utilizing nonpolyelectrolytes to modify cellulose, the interactions that dominate are hydrogen bonding and van der Waals forces. Previous research into the use of nonpolyelectrolytes to modify cellulose has taken inspiration from the components naturally found in conjunction with cellulose in plant cell walls, in particular, hemicelluloses. Hemicelluloses are a heterogeneous class of polysaccharides which, in nature, interact with cellulose and contribute to strengthening the cell wall.44 One interesting hemicellulose, which is well-known to bind to cellulose through noncovalent interactions, is xyloglucan (XG).45 XG is a branched polysaccharide based on the same anhydro-D-glucose backbone as cellulose, with xylose and galactose residues branching from the backbone; in some cases, fucose is also present depending on the XG source (for the XG structure, see SI, Figure S1).44 Laine and coworkers compared the adsorption of XG, galactoglucomannan (GGM), and several arabinoxylans from various sources to nanocellulose films by QCM-D and showed that XG had the highest amount of, and the most irreversible binding.46 A separate study by Gu and Catchmark showed similar results when comparing the adsorption of XG, xylan, arabinogalactan, and pectin. Generally, XG exhibited the highest binding affinity.47 Zhou et al. reviewed the use of XG in cellulose modification in 2007.48 The majority of cases discussed in the review utilize XG, which had been chemically modified with small molecules or enzymatically degraded into xyloglucan oligomers (XGO). It would be of great interest to the field if XG could be utilized as a molecular anchor to the cellulose surface while having it attached to a polymer, that is, as a block copolymer. Several block copolymers of XGO have been prepared,49−52 as have XG graft copolymers.53−56 However, these materials have not been evaluated in their adsorption to cellulose. Zhou et al. reported a modified XG that behaved as a molecular anchor to the cellulose surface for a grafting-from polymerization by atom transfer radical polymerization (ATRP).57 Further research from the same group showed the synthesis of a triblock copolymer XGO-b-poly(ethylene oxide)b-polystyrene (PS), which was found to adsorb to cellulose nanocrystals (CNC).49 The use of XG in cellulose-based nanocomposites has also been recently exemplified by Dammak et al., who reported the formation of XG-CNC multilayered thin films via layer-by-layer self-assembly.58 Therefore, XG is an attractive alternative to using polyelectrolytes for the modification of cellulose via physical adsorption, due to the excellent affinity between XG and cellulose, and importantly, as it is a naturally occurring biopolymer, it is a renewable, biodegradable resource. We thus anticipated that designing latex particles decorated by noncharged XG segments could be achieved through a PISA process, which required the modification of the chain end of a high molar mass XG17K (17 kg mol−1) with a chain transfer agent for RAFT. This transformation was performed using reducing end chemistry. The resulting XG17K-RAFT agent was utilized in the RAFT-mediated surfactant-free emulsion polymerization of MMA in order to form latex particles through PISA, consisting of a hydrophobic PMMA core sterically stabilized by the hydrophilic XG chains. The latexes of varying compositions were characterized and their physical adsorption to two different cellulosic substrates was investigated. Filter papers were used to assess the latex adsorption to a neutral cellulosic substrate while the adsorption of the particles to anionic model cellulose surfaces was also studied in situ by QCM-D. This novel procedure of employing a B
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
an EcoSEC RI detector and three columns (PSS PFG 5 μm; Microguard, 100 and 300 Å) with 0.01 M LiBr in DMF as the mobile phase at 50 °C and a flow rate of 0.2 mL min−1. A conventional calibration method was employed using PMMA standards and toluene was used as the internal flow marker. UV−Visible Spectroscopy. UV−visible spectra were recorded using a Shimadzu UV2550 spectrophotometer using quartz cells. Dynamic Light Scattering (DLS). Measurements were performed with a Malvern Zetasizer NanoZS at 25 °C. The z-average diameter (Dz) and the polydispersity index (PdI) of samples in Milli-Q water (1 mg mL−1 unless stated otherwise) were measured by DLS. The same instrument was also used for measuring the electrophoretic mobility in order to estimate the zeta potential (ζ) of the formed latexes. Transmission Electron Microscopy (TEM). Thin liquid films of the latex suspension were deposited onto 300 mesh Cu grids coated with a holey carbon film (Quantifoil R2/1; Agar Scientific, U.K.) and quenchfrozen in liquid ethane using a cryo-plunge workstation (made at LPS Orsay). The specimens were then mounted on a precooled Gatan 626 specimen holder, transferred into the Philips CM120 microscope operating at an accelerating voltage of 120 kV (Centre Technologique des Microstructures (CTμ), platform of the Claude Bernard Lyon 1 University, Villeurbanne, France). Scanning Electron Microscopy (SEM). A Hitachi S-4800 field emission scanning electron microscope was used to characterize the filter papers after adsorption of latex particles. Samples were sputter coated with a 5 nm layer of Pd−Au (Cressington sputter coater 208RH). Thermogravimetric Analysis (TGA). A TA Instruments Hi-Res TGA 2950 analyzer was used, operating in a N2 flow of 30 mL min−1, a heating rate of 10 °C min−1, heating the samples from 40 to 700 °C. Differential Scanning Calorimetry (DSC). Measurements were performed with a DSC 1 from Mettler-Toledo. Samples were measured through a cycle of heating−cooling−heating, with a starting temperature of −50 °C, up to 150 °C, at a heating/cooling rate of 10 °C min−1, under a N2 flow of 30 mL min−1. Tg values reported were taken from the second heating step. Contact Angle (CA). Contact angles were measured at 50% RH and 23 °C on a KSV instrument CAM 200 equipped with a Basler A602f camera, using 5 μL droplets of Milli-Q water. A Young−Laplace fitting model was used to process the images. Contact angle values reported were those observed after 30 s of measurement. Quartz Crystal Microbalance with Dissipation Monitoring (QCMD). Adsorption of the samples to model cellulose surfaces was studied using a QCM-E4 from Q-sense AB with a continuous flow of 0.15 mL min−1. Each sample was diluted with Milli-Q water (0.1 g L−1) for the adsorption experiments. This instrument measures the change in resonance frequency of the crystal, corresponding to a change in mass attached to the surface. To convert a change in frequency to its corresponding change in adsorbed mass per area unit, the Sauerbrey model63 was used:
hemicellulose as a macroRAFT in the PISA process to form XG decorated latex particles introduces a new platform for the biomimetic modification of various cellulosic substrates, including those such as cotton and bacterial cellulose, which inherently do not contain charges.
■
EXPERIMENTAL SECTION
Materials. Xyloglucan (XG) obtained from Tamarind seed (Tamarindus indica), which was enzymatically degraded (to an average Mn ≈ 17 kg mol−1, XG17K) consisting of xyloglucan oligomers XXXG, XXLG, XLXG, and XLLG in a ratio of 1.4:3:1:5.4,59,60 (SI, Figure S1) was kindly supplied by CelluTech AB, Sweden. Acetic acid (≥99%), 1ethyl-3-(3-(dimethylamino)propyl)carbodiimide (EDC, ≥97%), Nhydroxysuccinimide (NHS, 98%), 4-cyano-4(phenylcarbonothioylthio)pentanoic acid (CTP, ≥97%), methyl methacrylate (MMA, 99%), 2,2′-azobis(2-methylpropionamidine)dihydrochloride (AIBA, 97%), 4-methylmorpholine N-oxide solution (NMMO, 50 wt % in H2O), methanol (≥99%), and DMSO (≥99.7%) were purchased from Aldrich, sodium cyanoborohydride (NaCNBH3, 95%), hexamethylenediamine (HMD, 99.5%), were purchased from Acros Organics. Poly(vinyl amine) (PVAm, Lupamin 5095, Mw = 45.0 kg mol−1) was kindly supplied by BASF, Germany, and dialyzed against deionized water and freeze-dried prior to its use in the preparation of the model cellulose surfaces. Deuterated solvents were purchased from Cambridge Isotope Laboratories. All materials were used as received unless stated otherwise. The dialysis membrane used was Spectra/Por 6 standard RC with a molecular weight cut off (MWCO) of 50 kg mol−1, purchased from Spectrum Laboratories and rinsed with deionized water before use. The QCM crystals used were AT-cut crystals (5 MHz resonance frequency) with an active surface of sputtered silica (50 nm thickness) supplied by Q-sense AB. The cellulose fibers (Domsjö Dissolving Plus; Domsjö Aditya Birla AB, Domsjö, Sweden), used for the preparation of cellulose model surfaces, were pretreated by carboxymethylation to obtain anionic carboxylic charges (350 μeq g−1) on the cellulose fibers prior to use according to a procedure developed and described by Wågberg et al.61 Model Cellulose Surfaces. The model cellulose surfaces were prepared as follows: the dried pulp was dissolved in NMMO (50 wt % in H2O) at 125 °C, to a concentration of 20 mg mL−1, and diluted to 4× the original volume with DMSO. This solution was then spincoated (Spin-coater KW-4A, Chemat Technology) for 15 s at 1500 rpm, followed by 30 s at 3500 rpm onto a QCM crystal (AT-cut quartz crystals with silicon oxide surface) that had previously been placed in an air plasma cleaner (PDC 002, Harrick Scientific Corporation) for 2 min, then dipped in PVAm (0.1 g L−1 in Milli-Q, pH 7.5) for 15 min, rinsed with Milli-Q, and blown dry with N2.62 Characterization. Nuclear Magnetic Resonance (NMR). 1H NMR spectra were recorded at room temperature with a Bruker Avance 400 MHz spectrometer, using D2O or d6-DMSO as the deuterated solvents. Fourier Transform Infrared Spectroscopy (FT-IR). A PerkinElmer Spectrum 2000 FT-IR equipped with a MKII Golden Gate, single reflection ATR System from Specac Ltd., (London, U.K.) was used. The ATR-crystal used was a MKII heated Diamond 45° ATR Top Plate. For each spectrum, 16 scans were recorded. The spectra were normalized to the region between 4000 and 3100 cm−1 unless stated otherwise. Size Exclusion Chromatography (SEC). Average molar masses (number-average molar mass, Mn, and weight-average molar mass, Mw) and molar-mass dispersity (ĐM = Mw/Mn) were determined by SEC with a mobile phase of DMSO or DMF. SEC measurements in DMSO were performed on a SECcurity 1260 (PSS, Mainz, Germany) equipped with refractive index (RI) detector (40 °C) and three columns (PSS GRAM; precolumn, 10 μm 100 Å and 10 μm 10000 Å), with DMSO + 0.5 w/w% LiBr as the mobile phase at 60 °C and a flow rate of 0.5 mL min−1. A conventional calibration method was employed using pullulan standards. SEC measurements in DMF were performed using a TOSOH EcoSEC HLC-8320 system equipped with
Δm = C
Δf n
where C is a sensitivity constant, − 0.177 (mg (m2 Hz)−1), Δf the change in resonance frequency (Hz), and n the overtone number. The dissipation is related to the viscoelastic properties of the adsorbed layer. A thin, rigid attached film is expected to yield a low change in dissipation. A more water-rich and mobile film is expected to yield a larger change in dissipation. The dissipation factor, D, is defined as
D=
Edissipated 2πEstored
where Edissipated is the energy dissipated during one oscillation period, and Estored is the energy stored in the oscillating system. This Sauerbrey model assumes rigidly attached layers, and the attached amount determined contains both polymer and other compounds coupled to the surface. Earlier work has shown that this model is also valid for C
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
Table 1. Experimental Conditions and Results for the Aqueous RAFT-Mediated Emulsion Polymerization of MMA with XG17KRAFT target compositiona (τ-XG-PMMAx)
convb (%)
MMA DPtheor
Mn theorc (kg mol−1)
Mnd (kg mol−1)
Mwd (kg mol−1)
ĐM d
XG17K-RAFT 10-XG17K-PMMA50 10-XG17K-PMMA100 10-XG17K-PMMA175 5-XG17K-PMMA100 5-XG17K-PMMA175 5-XG17K-PMMA250 5-XG17K-PMMA375 5-XG17K-PMMA500
− 63 82 78 64 65 65 56 76
− 31 82 136 64 114 163 210 378
− 14.5 19.6 25.0 17.8 22.8 27.7 32.4 49.2
11.4 12.0 12.6 11.6 12.8 13.2 13.8
32.5 33.3 36.1 38.9 35.9 41.2 45.8
2.85 2.77 2.87 3.35 2.80 3.11 3.32
g
g
g
g
g
g
Mne (kg mol−1)
Mwe (kg mol−1)
ĐM e
−
−
−
g
g
g
53.2 72.5 58.3 61.8 67.2 81.5 96.0
73.7 125 85.7 129 103 116 128
1.39 1.72 1.47 2.08 1.54 1.43 1.33
Dz (nm)
PdI
− 152 152 275 125 145 154 160 545f
− 0.13 0.09 0.43 0.05 0.06 0.03 0.04 0.49f
τ = solids content, x = targeted DP. Calculated by gravimetry. Taking into account the conversion and using the experimental Mn of the XG17KRAFT by SEC (11.4 kg mol−1). dSEC with a mobile phase of DMSO + 0.5 w/w% LiBr using conventional calibration with pullulan standards. eSEC with a mobile phase of DMF + 0.01 M LiBr using conventional calibration with PMMA standards. fToo polydisperse for an accurate DLS measurement. gSample would not dissolve for SEC analysis.
a
b
c
samples purified by dialysis were used for the subsequent adsorption studies. Filter Paper Adsorption Studies. The filter papers (cut to 20 × 20 mm2) were immersed in 5.0 mL of the latex dispersions that had been diluted to a concentration of 5.0 mg mL−1 in Milli-Q water. The samples were placed on a shaking device for 150 min at 100 rpm for 2.5 h. A control was performed with no latex present and immersed in pure Milli-Q water. After the adsorption, the samples were rinsed with Milli-Q water and allowed to dry at room temperature. Annealing was conducted by placing the filter papers in glass vials in an oven at 160 °C for 1 h.
layers with higher dissipations and comparable to more advanced models.64 Atomic Force Microscopy (AFM). A Multimode 8 (Bruker, U.S.A.) was used with the ScanAsyst in Air mode, using a cantilever with 70 kHz resonance frequency, spring constant 0.4 N m−1, and tip radius 2 nm (ScanAsyst-Air, Bruker, U.S.A.). Samples for imaging were used directly after QCM-D experiments, after drying under a N2 flow or after an annealing step (160 °C for 1 h). Experimental Procedures. XG17K-NH2 synthesis. XG17K-NH2 was obtained as previously described.65,66 XG17K (20.0 g, 1.18 mmol) was dissolved in deionized water (200 mL), and the pH was adjusted to pH 5 using acetic acid. HMD (1.34 g, 11.8 mmol) was dissolved in deionized water (10.0 mL) in a round-bottom flask with stirring. The XG17K solution was added to the HMD solution slowly, and the pH was checked and again adjusted to pH 5 with acetic acid. NaCNBH3 (0.19 g, 2.9 mmol) was added and the reaction mixture was heated to 55 °C overnight (18 h). At this time more NaCNBH3 was added (0.19 g, 2.9 mmol) and the reaction was allowed to proceed for a further 6 h to ensure complete reaction. The product was obtained by precipitation directly into MeOH and dried in a vacuum oven at 50 °C to give a pale yellow solid, XG17K-NH2, 15.34 g (76%). The resulting product was analyzed by 1H NMR, FT-IR spectroscopy, and SEC; see SI. XG17K-RAFT Synthesis. XG17K-NH2 (14.7 g, 0.86 mmol) was dissolved in deionized water (150 mL). A solution of EDC (0.46 mL, 2.6 mmol) and NHS (0.30 g, 2.6 mmol) in deionized water (10.0 mL) was prepared and added to the XG17K-NH2 solution. CTP (0.60 g, 2.1 mmol) was added to the reaction mixture and the reaction was allowed to proceed at room temperature overnight. After 24 h, the product was obtained by precipitation into methanol and dried in a vacuum oven at 50 °C to yield a pale pink solid, XG17K-RAFT, 13.8 g (93%). The resulting product was analyzed by 1H NMR, FT-IR, UV− visible spectroscopy, and SEC; see SI. Exemplified Emulsion Polymerization (see Table 1, 10-XG17KPMMA50). XG17K-RAFT (0.75 g, 0.04 mmol) was dissolved in deionized water (22.0 mL). A solution of AIBA dissolved in deionized water was prepared (2.0 mg mL−1) of which 0.5 mL was added to the reaction mixture, followed by MMA (0.46 mL, 4.3 mmol). The reaction mixture was degassed with argon in an ice/water bath for 30 min. After degassing, the reaction vessel was sealed and placed in an oil bath set at 70 °C for 4 h. Samples were taken at certain time intervals for gravimetric, particle size and molar mass analyses. The reactions were terminated after 4 h by placing the reaction vessel in an ice bath. The pH of the reaction mixtures was determined to be close to pH 6, before and after polymerization. The degree of polymerization was controlled by varying the mass of XG17K-RAFT and monomer added while maintaining the desired solids content. A total of 5 mL of the crude reaction mixtures were purified by dialysis against deionized water using a 50 kg mol−1 molar mass cut off dialysis membrane. The
■
RESULTS AND DISCUSSION This work describes the facile preparation of XG17K-PMMAx latex particles via RAFT-mediated surfactant-free emulsion polymerization, designed for the modification of cellulose, negating the need for electrostatic interactions. Synthesis of Xyloglucan MacroRAFT Agent, XG17KRAFT. The reductive amination of the reducing chain end of xyloglucan has previously been reported utilizing facile reaction conditions.65,66 Here, through the reductive amination of the reducing chain end with an excess of HMD in water, the xyloglucan (Mn ≈ 17 kg mol−1, XG17K) was functionalized with an amino group, see Scheme 1, to yield XG17K-NH2. This allowed for further selective reaction of a carboxylic acid functional RAFT agent, CTP, with the primary amine through Scheme 1. Reaction Scheme for the Synthesis of XG17KRAFT Used in the Study
D
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
consistent with those described for the reaction performed at 5% solids content with XG17K-RAFT, targeting a DP of MMA of 175 monomer units (5-XG17K-PMMA175, Table 1). These blank experiments were both unsuccessful as macroscopic phase separation and precipitation were observed, whereas when the XG17K-RAFT was utilized, no precipitation was observed, see SI, Figure S6, and stable nanoparticles were formed. Therefore, it was concluded that the dithiobenzoate moiety was indeed necessary at the XG reducing chain end for stable particle formation. The possible degradation of the dithiobenzoate group was followed by UV−vis spectroscopy; confirming that the RAFT moiety was still intact after 4 h of heating at 70 °C at pH 6; therefore, minimal degradation of the dithiobenzoate would occur during the polymerizations, see SI, Figure S7. Various experiments were carried out with XG17K-RAFT targeting different DP of MMA for either 5 or 10% solids content (Table 1). The conversion over time was monitored by gravimetric analysis, and in each case, the polymerization had reached a plateau by 4 h reaction time. Figure 2A highlights typical kinetic plots for the emulsion polymerizations (for the kinetic evaluation of the other XG17K-PMMAx latex particles see SI, Figure S8). All the experiments show a similar profile: a small increase in conversion during the initial stage of the polymerization due to particle nucleation, a sharp increase in conversion which corresponds to propagation of the polymer in the core of the particles, followed by a plateau. While this plateau was reached the polymerizations did not reach full conversion, this was considered to be due to evaporation of MMA during the degassing step therefore gravimetric analysis would be inaccurate. The formation of particles throughout the polymerizations was followed by dynamic light scattering (DLS) to give the zaverage diameter (Dz) and polydispersity index (PdI) of the particles, see Figures 2B and S9. In our case, the particle formation was observed within 2−11% conversion. With further polymerization, the Dz values remained similar; however, the PdI values decreased with increasing conversions. Characterization of the final latex samples revealed particles in the range of 125−275 nm (Table 1). In the case of 10-XG17KPMMA175, both the size and PdI were unusually high, and sample 5-XG17K-PMMA500 was too polydisperse for an accurate DLS measurement, suggesting that the particles were unstable and aggregates had formed. Indeed, we found that in this system, increasing the solids content, τ, above 10%, or increasing the DP of the hydrophobic polymer led to aggregation of particles resulting in the precipitation of the formed polymer rather than the formation of stable particles. One explanation for this could be that, as the MMA content is increased the hydrophobic core of the particles increases in diameter and the stabilizing XG chains at the particle surface remain unchanged, the number of stabilizing chains per surface area of the particle decreases with increasing the MMA content. This could cause decreased colloidal stability, and therefore, a balance between the amounts of stabilizing hydrophilic XG and hydrophobic PMMA must be adhered to for ensuring successful particle stabilization by the XG17K-RAFT used in this system. The Dz values of the stable latexes were larger than what could be expected from the self-assembly of XG17K-b-PMMAx block copolymers and likely indicate that, despite its involvement in the particle stabilization by covalent anchoring, the XG17K-RAFT macroRAFT was unable to fully control the
EDC and NHS aqueous coupling to introduce a chain transfer agent moiety at the XG17K-NH2 chain end (Scheme 1). It is worth noting that while the primary amine may react with the dithiobenzoate group of the CTP, under these conditions, the reaction of the amine with the activated carboxylic acid was favored as the dithiobenzoate moiety remained intact, as previously reported by D’Agosto and co-workers.67 The successful coupling was confirmed by 1H NMR (see SI, Figure S2); the spectrum of XG17K-NH2 shows new chemical shifts, at 2.75 ppm and 1.6−1.2 ppm, corresponding to the −CH2 groups introduced. The presence of the aromatic group in the dithiobenzoate group in XG17K-RAFT is confirmed by signals at 7.7−7.3 ppm. Chemical shifts corresponding to other protons introduced were not observable due to the broad peaks attributed to XG (between 5.2 and 3.2 ppm). SEC analysis after each modification step gave similar molar masses, indicating that no coupling had occurred between two XG chains. UV− visible spectroscopy also confirmed the structure of the XG17KRAFT, showing an adsorption at 283 nm, attributed to the C S bond. See the SI for the 1H NMR, SEC, FT-IR, and UV−vis analyses of XG17K-NH2 and XG17K-RAFT, Figures S2−5 and Table S1. Synthesis of XG17K-PMMAx Latex Particles. Due to the desire to modify cellulose with hydrophobic polymers and the limited solubility of xyloglucan, dissolving in only water or DMSO, solvents not compatible with many hydrophobic monomers and polymers, the chain extension of XG17KRAFT was investigated in aqueous emulsion polymerization. Building on the work previously conducted in our group,39 the polymerization of MMA in a RAFT-mediated surfactant-free emulsion polymerization system was conducted, Figure 1. Various polymerization degrees (DP) for the hydrophobic block were targeted while maintaining the total solids content at 5 and 10 wt %, see Table 1.
Figure 1. Schematic representation of the formation of XG17K-PMMAx latex particles.
The polymerizations were conducted at 70 °C, employing AIBA as the radical initiator, over a period of 4 h at pH 6. To confirm the activity of the novel XG17K-RAFT as a suitable macroRAFT in the PISA process, blank experiments were performed where XG17K-RAFT was substituted with either XG17K or XG17K-NH2. The reaction conditions used were E
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
Figure 2. Kinetic evaluation with (A) conversion and (B) z-average diameter (closed symbols) and PdI (open symbols) of samples: 10-XG17KPMMA100 (black squares), 5-XG17K-PMMA100 (red triangles), 5-XG17K-PMMA175 (blue circles), and 5-XG17K-PMMA250 (green diamonds).
Figure 3. TEM images of the final latex samples before dialysis: (A) 5-XG17K-PMMA100, (B) 5-XG17K-PMMA175, (C) 5-XG17K-PMMA250, and (D) 5XG17K-PMMA375; Scale bars 100 nm.
lower the size (see the 5% solids content series). Cryo-TEM analyses (Figure 3) of the 5-XG17K-PMMA100−375 samples showed that while the particles appeared spherical in all cases, they displayed a rather rough, uneven particle surface. This, and the previously mentioned limitations to the system, could be a result of the XG structure, which, as previously reported, has a stiff, extended conformation68 and tendency to aggregate in solution.69 To summarize, the XG17K-RAFT was proven to be necessary to ensure stabilization of the particles; however, it was not a particularly robust stabilizer when pushing the reaction conditions to higher solids content or hydrophobic polymer content. The latex samples were dialyzed to remove any unreacted XG17K-RAFT that could be present in the samples. Dialysis was performed against deionized water using a dialysis membrane with a molar mass cut off of 50 kg mol−1 to ensure that only unreacted XG17K-RAFT and small molecules could cross the membrane. The first removal of water from the dialysis vessel was freeze-dried to assess how much, if any, XG17K-RAFT was removed from the latex samples. In all cases, the quantity was too low to be accurately determined (
Xyloglucan-Functional Latex Particles via RAFT-Mediated Emulsion Polymerization for the Biomimetic Modification of Cellulose Fiona L. Hatton,† Marcus Ruda,‡ Muriel Lansalot,§ Franck D’Agosto,§ Eva Malmström,† and Anna Carlmark*,† †
KTH Royal Institute of Technology, School of Chemical Science and Engineering, Department of Fibre and Polymer Technology, Teknikringen 56, SE-100 44 Stockholm, Sweden ‡ CelluTech AB, Teknikringen 38, SE-114 28 Stockholm, Sweden § Université de Lyon, Univ Lyon 1, CPE Lyon, CNRS, UMR 5265, C2P2 (Chemistry, Catalysis, Polymers and Processes), Team LCPP, Bat 308F, 43 Bd du 11 Novembre 1918, 69616 Villeurbanne, France S Supporting Information *
ABSTRACT: Herein, we report a novel class of latex particles composed of a hemicellulose, xyloglucan (XG), and poly(methyl methacrylate) (PMMA), specially designed to enable a biomimetic modification of cellulose. The formation of the latex particles was achieved utilizing reversible addition− fragmentation chain transfer (RAFT) mediated surfactant-free emulsion polymerization employing XG as a hydrophilic macromolecular RAFT agent (macroRAFT). In an initial step, XG was functionalized at the reducing chain end to bear a dithioester. This XG macroRAFT was subsequently utilized in water and chain extended with methyl methacrylate (MMA) as hydrophobic monomer, inspired by a polymerization-induced self-assembly (PISA) process. This yielded latex nanoparticles with a hydrophobic PMMA core stabilized by the hydrophilic XG chains at the corona. The molar mass of PMMA targeted was varied, resulting in a series of stable latex particles with hydrophobic PMMA content between 22 and 68 wt % of the total solids content (5−10%). The XG-PMMA nanoparticles were subsequently adsorbed to a neutral cellulose substrate (filter paper), and the modified surfaces were analyzed by FT-IR and SEM analyses. The adsorption of the latex particles was also investigated by quartz crystal microbalance with dissipation monitoring (QCM-D), where the nanoparticles were adsorbed to negatively charged model cellulose surfaces. The surfaces were analyzed by atomic force microscopy (AFM) and contact angle (CA) measurements. QCMD experiments showed that more mass was adsorbed to the surfaces with increasing molar mass of the PMMA present. AFM of the surfaces after adsorption showed discrete particles, which were no longer present after annealing (160 °C, 1 h) and the roughness (Rq) of the surfaces had also decreased by at least half. Interestingly, after annealing, the surfaces did not all become more hydrophobic, as monitored by CA measurements, indicating that the surface roughness was an important factor to consider when evaluating the surface properties following particle adsorption. This novel class of latex nanoparticles provides an excellent platform for cellulose modification via physical adsorption. The utilization of XG as the anchoring molecule to cellulose provides a versatile methodology, as it does not rely on electrostatic interactions for the physical adsorption, enabling a wide range of cellulose substrates to be modified, including neutral sources such as cotton and bacterial nanocellulose, leading to new and advanced materials.
■
functionalities.2 Modification of cellulose with polymers can be achieved by covalent attachment via grafting-from or grafting-to methods5−7 or by physical adsorption exploiting electrostatic interactions, hydrogen bonding, or van der Waals forces.8 One of the largest industries based around cellulose, the pulp and paper industry, has utilized polyelectrolytes for decades to modify cellulose fibers via physical adsorption. This physical adsorption, or physisorption, is achieved through electrostatic
INTRODUCTION
The incorporation of cellulose in biocomposites has been a research area of significant interest in recent years due to the natural abundance, physical properties and biorenewability of cellulose, the main constituent of plant cell walls.1 The growing demand for biobased materials to replace fossil fuel-derived materials has most likely stimulated this interest into biocomposite research.2−4 However, due to the hydrophilic nature of cellulose, it has poor compatibility with hydrophobic matrix polymers commonly used in composites. Therefore, the modification of cellulose is necessary to impart desired properties, such as improving miscibility with hydrophobic polymers, barrier properties and incorporating specific © XXXX American Chemical Society
Received: January 11, 2016 Revised: February 24, 2016
A
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules interactions between a cationic polyelectrolyte and anionic cellulose substrates.9 The introduction of negative charges to cellulose is accomplished through oxidation of the primary hydroxy groups present on each repeating unit (anhydro-Dglucose) of the cellulose chain, for example, by carboxylation or sulfonation, therefore, enabling electrostatic interactions to occur with cationic species.1 Previous work in our group has focused on the modification of cellulose with polymers both through covalent10−13 and electrostatic14−17 interactions. Targeting physical adsorption through electrostatic interactions, tailor-made block copolymers have been utilized, which contain a quaternized poly(dimethylaminoethyl methacrylate) (PDMAEMA) block14−16 or a poly((acrylamidopropyl) trimethylammonium chloride) (PAPTAC) block.17 Cationic latexes are well-known for the modification of cellulose. Studies into the reinforcement of paper have been reported utilizing cationic latexes of poly(styrene-co-butadiene).18−20 Cationic latexes composed of poly(styrene-co-nbutyl acrylate) were used to prepare nanocomposite films with cellulose whiskers to improve mechanical properties.21 The same type of latexes have been studied for their adhesive properties to cellulose22 and their adsorption to cotton fibers.23 Core−shell latex particles of varying compositions of styrene, nbutyl acrylate, methyl methacrylate, and 2-ethylhexyl acrylate were also investigated in their adsorption to cellulose fibers, increasing the hydrophobicity and enabling incorporation into a polypropylene-based biocomposite.24−26 The general principle is always to design latex particles carrying cationic charges at the surface that can interact with the negatively charged cellulose and a hydrophobic core polymer of interest for the targeted cellulose modification. First described by Hawkett and co-workers in 2005,27 polymerization-induced self-assembly (PISA) describes the chain extension, in water, of hydrophilic polymer chains, synthesized by controlled radical polymerization, with a hydrophobic monomer.28−31 When the growth of the hydrophobic segment is well-controlled, the resulting latex nanoparticles are exclusively composed of amphiphilic block copolymers. The PISA process has been well described previously, including the one-pot synthesis of latex particles in water.32−34 While not only an excellent tool to design amphiphilic block copolymers in high yield directly in water, PISA is also a very efficient method to design the particle surface at will. For example, a variety of original, carbohydrate functional nanoparticles have already been obtained,35−37 including alginate-decorated nanoparticles.38 With the aim of modifying the surface of cellulose, we have recently reported the synthesis of poly(methyl methacrylate) (PMMA) latex particles, utilizing the PISA mechanism, stabilized by P(DMAEMA-co-methacrylic acid (MAA)) macroRAFT segments.39 The manipulation of pH led to the protonation of the DMAEMA repeating units, resulting in cationic latexes which were subsequently investigated for their adsorption to a negatively charged cellulose substrate as followed by quartz crystal microbalance with dissipation monitoring (QCM-D). As mentioned previously, the use of cationic polyelectrolytes and latexes to modify cellulose has been well studied,9,40 including adsorption to cellulosic substrates such as nanocelluloses.14,41,42 Although the use of nonpolyelectrolytes, or neutral polymers, to modify cellulose has not been as well investigated, Perrier and co-workers have studied the adsorption of synthetic polymers with varying architectures to cellulose, including glucose units for cellulose recognition.43
When utilizing nonpolyelectrolytes to modify cellulose, the interactions that dominate are hydrogen bonding and van der Waals forces. Previous research into the use of nonpolyelectrolytes to modify cellulose has taken inspiration from the components naturally found in conjunction with cellulose in plant cell walls, in particular, hemicelluloses. Hemicelluloses are a heterogeneous class of polysaccharides which, in nature, interact with cellulose and contribute to strengthening the cell wall.44 One interesting hemicellulose, which is well-known to bind to cellulose through noncovalent interactions, is xyloglucan (XG).45 XG is a branched polysaccharide based on the same anhydro-D-glucose backbone as cellulose, with xylose and galactose residues branching from the backbone; in some cases, fucose is also present depending on the XG source (for the XG structure, see SI, Figure S1).44 Laine and coworkers compared the adsorption of XG, galactoglucomannan (GGM), and several arabinoxylans from various sources to nanocellulose films by QCM-D and showed that XG had the highest amount of, and the most irreversible binding.46 A separate study by Gu and Catchmark showed similar results when comparing the adsorption of XG, xylan, arabinogalactan, and pectin. Generally, XG exhibited the highest binding affinity.47 Zhou et al. reviewed the use of XG in cellulose modification in 2007.48 The majority of cases discussed in the review utilize XG, which had been chemically modified with small molecules or enzymatically degraded into xyloglucan oligomers (XGO). It would be of great interest to the field if XG could be utilized as a molecular anchor to the cellulose surface while having it attached to a polymer, that is, as a block copolymer. Several block copolymers of XGO have been prepared,49−52 as have XG graft copolymers.53−56 However, these materials have not been evaluated in their adsorption to cellulose. Zhou et al. reported a modified XG that behaved as a molecular anchor to the cellulose surface for a grafting-from polymerization by atom transfer radical polymerization (ATRP).57 Further research from the same group showed the synthesis of a triblock copolymer XGO-b-poly(ethylene oxide)b-polystyrene (PS), which was found to adsorb to cellulose nanocrystals (CNC).49 The use of XG in cellulose-based nanocomposites has also been recently exemplified by Dammak et al., who reported the formation of XG-CNC multilayered thin films via layer-by-layer self-assembly.58 Therefore, XG is an attractive alternative to using polyelectrolytes for the modification of cellulose via physical adsorption, due to the excellent affinity between XG and cellulose, and importantly, as it is a naturally occurring biopolymer, it is a renewable, biodegradable resource. We thus anticipated that designing latex particles decorated by noncharged XG segments could be achieved through a PISA process, which required the modification of the chain end of a high molar mass XG17K (17 kg mol−1) with a chain transfer agent for RAFT. This transformation was performed using reducing end chemistry. The resulting XG17K-RAFT agent was utilized in the RAFT-mediated surfactant-free emulsion polymerization of MMA in order to form latex particles through PISA, consisting of a hydrophobic PMMA core sterically stabilized by the hydrophilic XG chains. The latexes of varying compositions were characterized and their physical adsorption to two different cellulosic substrates was investigated. Filter papers were used to assess the latex adsorption to a neutral cellulosic substrate while the adsorption of the particles to anionic model cellulose surfaces was also studied in situ by QCM-D. This novel procedure of employing a B
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
an EcoSEC RI detector and three columns (PSS PFG 5 μm; Microguard, 100 and 300 Å) with 0.01 M LiBr in DMF as the mobile phase at 50 °C and a flow rate of 0.2 mL min−1. A conventional calibration method was employed using PMMA standards and toluene was used as the internal flow marker. UV−Visible Spectroscopy. UV−visible spectra were recorded using a Shimadzu UV2550 spectrophotometer using quartz cells. Dynamic Light Scattering (DLS). Measurements were performed with a Malvern Zetasizer NanoZS at 25 °C. The z-average diameter (Dz) and the polydispersity index (PdI) of samples in Milli-Q water (1 mg mL−1 unless stated otherwise) were measured by DLS. The same instrument was also used for measuring the electrophoretic mobility in order to estimate the zeta potential (ζ) of the formed latexes. Transmission Electron Microscopy (TEM). Thin liquid films of the latex suspension were deposited onto 300 mesh Cu grids coated with a holey carbon film (Quantifoil R2/1; Agar Scientific, U.K.) and quenchfrozen in liquid ethane using a cryo-plunge workstation (made at LPS Orsay). The specimens were then mounted on a precooled Gatan 626 specimen holder, transferred into the Philips CM120 microscope operating at an accelerating voltage of 120 kV (Centre Technologique des Microstructures (CTμ), platform of the Claude Bernard Lyon 1 University, Villeurbanne, France). Scanning Electron Microscopy (SEM). A Hitachi S-4800 field emission scanning electron microscope was used to characterize the filter papers after adsorption of latex particles. Samples were sputter coated with a 5 nm layer of Pd−Au (Cressington sputter coater 208RH). Thermogravimetric Analysis (TGA). A TA Instruments Hi-Res TGA 2950 analyzer was used, operating in a N2 flow of 30 mL min−1, a heating rate of 10 °C min−1, heating the samples from 40 to 700 °C. Differential Scanning Calorimetry (DSC). Measurements were performed with a DSC 1 from Mettler-Toledo. Samples were measured through a cycle of heating−cooling−heating, with a starting temperature of −50 °C, up to 150 °C, at a heating/cooling rate of 10 °C min−1, under a N2 flow of 30 mL min−1. Tg values reported were taken from the second heating step. Contact Angle (CA). Contact angles were measured at 50% RH and 23 °C on a KSV instrument CAM 200 equipped with a Basler A602f camera, using 5 μL droplets of Milli-Q water. A Young−Laplace fitting model was used to process the images. Contact angle values reported were those observed after 30 s of measurement. Quartz Crystal Microbalance with Dissipation Monitoring (QCMD). Adsorption of the samples to model cellulose surfaces was studied using a QCM-E4 from Q-sense AB with a continuous flow of 0.15 mL min−1. Each sample was diluted with Milli-Q water (0.1 g L−1) for the adsorption experiments. This instrument measures the change in resonance frequency of the crystal, corresponding to a change in mass attached to the surface. To convert a change in frequency to its corresponding change in adsorbed mass per area unit, the Sauerbrey model63 was used:
hemicellulose as a macroRAFT in the PISA process to form XG decorated latex particles introduces a new platform for the biomimetic modification of various cellulosic substrates, including those such as cotton and bacterial cellulose, which inherently do not contain charges.
■
EXPERIMENTAL SECTION
Materials. Xyloglucan (XG) obtained from Tamarind seed (Tamarindus indica), which was enzymatically degraded (to an average Mn ≈ 17 kg mol−1, XG17K) consisting of xyloglucan oligomers XXXG, XXLG, XLXG, and XLLG in a ratio of 1.4:3:1:5.4,59,60 (SI, Figure S1) was kindly supplied by CelluTech AB, Sweden. Acetic acid (≥99%), 1ethyl-3-(3-(dimethylamino)propyl)carbodiimide (EDC, ≥97%), Nhydroxysuccinimide (NHS, 98%), 4-cyano-4(phenylcarbonothioylthio)pentanoic acid (CTP, ≥97%), methyl methacrylate (MMA, 99%), 2,2′-azobis(2-methylpropionamidine)dihydrochloride (AIBA, 97%), 4-methylmorpholine N-oxide solution (NMMO, 50 wt % in H2O), methanol (≥99%), and DMSO (≥99.7%) were purchased from Aldrich, sodium cyanoborohydride (NaCNBH3, 95%), hexamethylenediamine (HMD, 99.5%), were purchased from Acros Organics. Poly(vinyl amine) (PVAm, Lupamin 5095, Mw = 45.0 kg mol−1) was kindly supplied by BASF, Germany, and dialyzed against deionized water and freeze-dried prior to its use in the preparation of the model cellulose surfaces. Deuterated solvents were purchased from Cambridge Isotope Laboratories. All materials were used as received unless stated otherwise. The dialysis membrane used was Spectra/Por 6 standard RC with a molecular weight cut off (MWCO) of 50 kg mol−1, purchased from Spectrum Laboratories and rinsed with deionized water before use. The QCM crystals used were AT-cut crystals (5 MHz resonance frequency) with an active surface of sputtered silica (50 nm thickness) supplied by Q-sense AB. The cellulose fibers (Domsjö Dissolving Plus; Domsjö Aditya Birla AB, Domsjö, Sweden), used for the preparation of cellulose model surfaces, were pretreated by carboxymethylation to obtain anionic carboxylic charges (350 μeq g−1) on the cellulose fibers prior to use according to a procedure developed and described by Wågberg et al.61 Model Cellulose Surfaces. The model cellulose surfaces were prepared as follows: the dried pulp was dissolved in NMMO (50 wt % in H2O) at 125 °C, to a concentration of 20 mg mL−1, and diluted to 4× the original volume with DMSO. This solution was then spincoated (Spin-coater KW-4A, Chemat Technology) for 15 s at 1500 rpm, followed by 30 s at 3500 rpm onto a QCM crystal (AT-cut quartz crystals with silicon oxide surface) that had previously been placed in an air plasma cleaner (PDC 002, Harrick Scientific Corporation) for 2 min, then dipped in PVAm (0.1 g L−1 in Milli-Q, pH 7.5) for 15 min, rinsed with Milli-Q, and blown dry with N2.62 Characterization. Nuclear Magnetic Resonance (NMR). 1H NMR spectra were recorded at room temperature with a Bruker Avance 400 MHz spectrometer, using D2O or d6-DMSO as the deuterated solvents. Fourier Transform Infrared Spectroscopy (FT-IR). A PerkinElmer Spectrum 2000 FT-IR equipped with a MKII Golden Gate, single reflection ATR System from Specac Ltd., (London, U.K.) was used. The ATR-crystal used was a MKII heated Diamond 45° ATR Top Plate. For each spectrum, 16 scans were recorded. The spectra were normalized to the region between 4000 and 3100 cm−1 unless stated otherwise. Size Exclusion Chromatography (SEC). Average molar masses (number-average molar mass, Mn, and weight-average molar mass, Mw) and molar-mass dispersity (ĐM = Mw/Mn) were determined by SEC with a mobile phase of DMSO or DMF. SEC measurements in DMSO were performed on a SECcurity 1260 (PSS, Mainz, Germany) equipped with refractive index (RI) detector (40 °C) and three columns (PSS GRAM; precolumn, 10 μm 100 Å and 10 μm 10000 Å), with DMSO + 0.5 w/w% LiBr as the mobile phase at 60 °C and a flow rate of 0.5 mL min−1. A conventional calibration method was employed using pullulan standards. SEC measurements in DMF were performed using a TOSOH EcoSEC HLC-8320 system equipped with
Δm = C
Δf n
where C is a sensitivity constant, − 0.177 (mg (m2 Hz)−1), Δf the change in resonance frequency (Hz), and n the overtone number. The dissipation is related to the viscoelastic properties of the adsorbed layer. A thin, rigid attached film is expected to yield a low change in dissipation. A more water-rich and mobile film is expected to yield a larger change in dissipation. The dissipation factor, D, is defined as
D=
Edissipated 2πEstored
where Edissipated is the energy dissipated during one oscillation period, and Estored is the energy stored in the oscillating system. This Sauerbrey model assumes rigidly attached layers, and the attached amount determined contains both polymer and other compounds coupled to the surface. Earlier work has shown that this model is also valid for C
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
Table 1. Experimental Conditions and Results for the Aqueous RAFT-Mediated Emulsion Polymerization of MMA with XG17KRAFT target compositiona (τ-XG-PMMAx)
convb (%)
MMA DPtheor
Mn theorc (kg mol−1)
Mnd (kg mol−1)
Mwd (kg mol−1)
ĐM d
XG17K-RAFT 10-XG17K-PMMA50 10-XG17K-PMMA100 10-XG17K-PMMA175 5-XG17K-PMMA100 5-XG17K-PMMA175 5-XG17K-PMMA250 5-XG17K-PMMA375 5-XG17K-PMMA500
− 63 82 78 64 65 65 56 76
− 31 82 136 64 114 163 210 378
− 14.5 19.6 25.0 17.8 22.8 27.7 32.4 49.2
11.4 12.0 12.6 11.6 12.8 13.2 13.8
32.5 33.3 36.1 38.9 35.9 41.2 45.8
2.85 2.77 2.87 3.35 2.80 3.11 3.32
g
g
g
g
g
g
Mne (kg mol−1)
Mwe (kg mol−1)
ĐM e
−
−
−
g
g
g
53.2 72.5 58.3 61.8 67.2 81.5 96.0
73.7 125 85.7 129 103 116 128
1.39 1.72 1.47 2.08 1.54 1.43 1.33
Dz (nm)
PdI
− 152 152 275 125 145 154 160 545f
− 0.13 0.09 0.43 0.05 0.06 0.03 0.04 0.49f
τ = solids content, x = targeted DP. Calculated by gravimetry. Taking into account the conversion and using the experimental Mn of the XG17KRAFT by SEC (11.4 kg mol−1). dSEC with a mobile phase of DMSO + 0.5 w/w% LiBr using conventional calibration with pullulan standards. eSEC with a mobile phase of DMF + 0.01 M LiBr using conventional calibration with PMMA standards. fToo polydisperse for an accurate DLS measurement. gSample would not dissolve for SEC analysis.
a
b
c
samples purified by dialysis were used for the subsequent adsorption studies. Filter Paper Adsorption Studies. The filter papers (cut to 20 × 20 mm2) were immersed in 5.0 mL of the latex dispersions that had been diluted to a concentration of 5.0 mg mL−1 in Milli-Q water. The samples were placed on a shaking device for 150 min at 100 rpm for 2.5 h. A control was performed with no latex present and immersed in pure Milli-Q water. After the adsorption, the samples were rinsed with Milli-Q water and allowed to dry at room temperature. Annealing was conducted by placing the filter papers in glass vials in an oven at 160 °C for 1 h.
layers with higher dissipations and comparable to more advanced models.64 Atomic Force Microscopy (AFM). A Multimode 8 (Bruker, U.S.A.) was used with the ScanAsyst in Air mode, using a cantilever with 70 kHz resonance frequency, spring constant 0.4 N m−1, and tip radius 2 nm (ScanAsyst-Air, Bruker, U.S.A.). Samples for imaging were used directly after QCM-D experiments, after drying under a N2 flow or after an annealing step (160 °C for 1 h). Experimental Procedures. XG17K-NH2 synthesis. XG17K-NH2 was obtained as previously described.65,66 XG17K (20.0 g, 1.18 mmol) was dissolved in deionized water (200 mL), and the pH was adjusted to pH 5 using acetic acid. HMD (1.34 g, 11.8 mmol) was dissolved in deionized water (10.0 mL) in a round-bottom flask with stirring. The XG17K solution was added to the HMD solution slowly, and the pH was checked and again adjusted to pH 5 with acetic acid. NaCNBH3 (0.19 g, 2.9 mmol) was added and the reaction mixture was heated to 55 °C overnight (18 h). At this time more NaCNBH3 was added (0.19 g, 2.9 mmol) and the reaction was allowed to proceed for a further 6 h to ensure complete reaction. The product was obtained by precipitation directly into MeOH and dried in a vacuum oven at 50 °C to give a pale yellow solid, XG17K-NH2, 15.34 g (76%). The resulting product was analyzed by 1H NMR, FT-IR spectroscopy, and SEC; see SI. XG17K-RAFT Synthesis. XG17K-NH2 (14.7 g, 0.86 mmol) was dissolved in deionized water (150 mL). A solution of EDC (0.46 mL, 2.6 mmol) and NHS (0.30 g, 2.6 mmol) in deionized water (10.0 mL) was prepared and added to the XG17K-NH2 solution. CTP (0.60 g, 2.1 mmol) was added to the reaction mixture and the reaction was allowed to proceed at room temperature overnight. After 24 h, the product was obtained by precipitation into methanol and dried in a vacuum oven at 50 °C to yield a pale pink solid, XG17K-RAFT, 13.8 g (93%). The resulting product was analyzed by 1H NMR, FT-IR, UV− visible spectroscopy, and SEC; see SI. Exemplified Emulsion Polymerization (see Table 1, 10-XG17KPMMA50). XG17K-RAFT (0.75 g, 0.04 mmol) was dissolved in deionized water (22.0 mL). A solution of AIBA dissolved in deionized water was prepared (2.0 mg mL−1) of which 0.5 mL was added to the reaction mixture, followed by MMA (0.46 mL, 4.3 mmol). The reaction mixture was degassed with argon in an ice/water bath for 30 min. After degassing, the reaction vessel was sealed and placed in an oil bath set at 70 °C for 4 h. Samples were taken at certain time intervals for gravimetric, particle size and molar mass analyses. The reactions were terminated after 4 h by placing the reaction vessel in an ice bath. The pH of the reaction mixtures was determined to be close to pH 6, before and after polymerization. The degree of polymerization was controlled by varying the mass of XG17K-RAFT and monomer added while maintaining the desired solids content. A total of 5 mL of the crude reaction mixtures were purified by dialysis against deionized water using a 50 kg mol−1 molar mass cut off dialysis membrane. The
■
RESULTS AND DISCUSSION This work describes the facile preparation of XG17K-PMMAx latex particles via RAFT-mediated surfactant-free emulsion polymerization, designed for the modification of cellulose, negating the need for electrostatic interactions. Synthesis of Xyloglucan MacroRAFT Agent, XG17KRAFT. The reductive amination of the reducing chain end of xyloglucan has previously been reported utilizing facile reaction conditions.65,66 Here, through the reductive amination of the reducing chain end with an excess of HMD in water, the xyloglucan (Mn ≈ 17 kg mol−1, XG17K) was functionalized with an amino group, see Scheme 1, to yield XG17K-NH2. This allowed for further selective reaction of a carboxylic acid functional RAFT agent, CTP, with the primary amine through Scheme 1. Reaction Scheme for the Synthesis of XG17KRAFT Used in the Study
D
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
consistent with those described for the reaction performed at 5% solids content with XG17K-RAFT, targeting a DP of MMA of 175 monomer units (5-XG17K-PMMA175, Table 1). These blank experiments were both unsuccessful as macroscopic phase separation and precipitation were observed, whereas when the XG17K-RAFT was utilized, no precipitation was observed, see SI, Figure S6, and stable nanoparticles were formed. Therefore, it was concluded that the dithiobenzoate moiety was indeed necessary at the XG reducing chain end for stable particle formation. The possible degradation of the dithiobenzoate group was followed by UV−vis spectroscopy; confirming that the RAFT moiety was still intact after 4 h of heating at 70 °C at pH 6; therefore, minimal degradation of the dithiobenzoate would occur during the polymerizations, see SI, Figure S7. Various experiments were carried out with XG17K-RAFT targeting different DP of MMA for either 5 or 10% solids content (Table 1). The conversion over time was monitored by gravimetric analysis, and in each case, the polymerization had reached a plateau by 4 h reaction time. Figure 2A highlights typical kinetic plots for the emulsion polymerizations (for the kinetic evaluation of the other XG17K-PMMAx latex particles see SI, Figure S8). All the experiments show a similar profile: a small increase in conversion during the initial stage of the polymerization due to particle nucleation, a sharp increase in conversion which corresponds to propagation of the polymer in the core of the particles, followed by a plateau. While this plateau was reached the polymerizations did not reach full conversion, this was considered to be due to evaporation of MMA during the degassing step therefore gravimetric analysis would be inaccurate. The formation of particles throughout the polymerizations was followed by dynamic light scattering (DLS) to give the zaverage diameter (Dz) and polydispersity index (PdI) of the particles, see Figures 2B and S9. In our case, the particle formation was observed within 2−11% conversion. With further polymerization, the Dz values remained similar; however, the PdI values decreased with increasing conversions. Characterization of the final latex samples revealed particles in the range of 125−275 nm (Table 1). In the case of 10-XG17KPMMA175, both the size and PdI were unusually high, and sample 5-XG17K-PMMA500 was too polydisperse for an accurate DLS measurement, suggesting that the particles were unstable and aggregates had formed. Indeed, we found that in this system, increasing the solids content, τ, above 10%, or increasing the DP of the hydrophobic polymer led to aggregation of particles resulting in the precipitation of the formed polymer rather than the formation of stable particles. One explanation for this could be that, as the MMA content is increased the hydrophobic core of the particles increases in diameter and the stabilizing XG chains at the particle surface remain unchanged, the number of stabilizing chains per surface area of the particle decreases with increasing the MMA content. This could cause decreased colloidal stability, and therefore, a balance between the amounts of stabilizing hydrophilic XG and hydrophobic PMMA must be adhered to for ensuring successful particle stabilization by the XG17K-RAFT used in this system. The Dz values of the stable latexes were larger than what could be expected from the self-assembly of XG17K-b-PMMAx block copolymers and likely indicate that, despite its involvement in the particle stabilization by covalent anchoring, the XG17K-RAFT macroRAFT was unable to fully control the
EDC and NHS aqueous coupling to introduce a chain transfer agent moiety at the XG17K-NH2 chain end (Scheme 1). It is worth noting that while the primary amine may react with the dithiobenzoate group of the CTP, under these conditions, the reaction of the amine with the activated carboxylic acid was favored as the dithiobenzoate moiety remained intact, as previously reported by D’Agosto and co-workers.67 The successful coupling was confirmed by 1H NMR (see SI, Figure S2); the spectrum of XG17K-NH2 shows new chemical shifts, at 2.75 ppm and 1.6−1.2 ppm, corresponding to the −CH2 groups introduced. The presence of the aromatic group in the dithiobenzoate group in XG17K-RAFT is confirmed by signals at 7.7−7.3 ppm. Chemical shifts corresponding to other protons introduced were not observable due to the broad peaks attributed to XG (between 5.2 and 3.2 ppm). SEC analysis after each modification step gave similar molar masses, indicating that no coupling had occurred between two XG chains. UV− visible spectroscopy also confirmed the structure of the XG17KRAFT, showing an adsorption at 283 nm, attributed to the C S bond. See the SI for the 1H NMR, SEC, FT-IR, and UV−vis analyses of XG17K-NH2 and XG17K-RAFT, Figures S2−5 and Table S1. Synthesis of XG17K-PMMAx Latex Particles. Due to the desire to modify cellulose with hydrophobic polymers and the limited solubility of xyloglucan, dissolving in only water or DMSO, solvents not compatible with many hydrophobic monomers and polymers, the chain extension of XG17KRAFT was investigated in aqueous emulsion polymerization. Building on the work previously conducted in our group,39 the polymerization of MMA in a RAFT-mediated surfactant-free emulsion polymerization system was conducted, Figure 1. Various polymerization degrees (DP) for the hydrophobic block were targeted while maintaining the total solids content at 5 and 10 wt %, see Table 1.
Figure 1. Schematic representation of the formation of XG17K-PMMAx latex particles.
The polymerizations were conducted at 70 °C, employing AIBA as the radical initiator, over a period of 4 h at pH 6. To confirm the activity of the novel XG17K-RAFT as a suitable macroRAFT in the PISA process, blank experiments were performed where XG17K-RAFT was substituted with either XG17K or XG17K-NH2. The reaction conditions used were E
DOI: 10.1021/acs.biomac.6b00036 Biomacromolecules XXXX, XXX, XXX−XXX
Article
Biomacromolecules
Figure 2. Kinetic evaluation with (A) conversion and (B) z-average diameter (closed symbols) and PdI (open symbols) of samples: 10-XG17KPMMA100 (black squares), 5-XG17K-PMMA100 (red triangles), 5-XG17K-PMMA175 (blue circles), and 5-XG17K-PMMA250 (green diamonds).
Figure 3. TEM images of the final latex samples before dialysis: (A) 5-XG17K-PMMA100, (B) 5-XG17K-PMMA175, (C) 5-XG17K-PMMA250, and (D) 5XG17K-PMMA375; Scale bars 100 nm.
lower the size (see the 5% solids content series). Cryo-TEM analyses (Figure 3) of the 5-XG17K-PMMA100−375 samples showed that while the particles appeared spherical in all cases, they displayed a rather rough, uneven particle surface. This, and the previously mentioned limitations to the system, could be a result of the XG structure, which, as previously reported, has a stiff, extended conformation68 and tendency to aggregate in solution.69 To summarize, the XG17K-RAFT was proven to be necessary to ensure stabilization of the particles; however, it was not a particularly robust stabilizer when pushing the reaction conditions to higher solids content or hydrophobic polymer content. The latex samples were dialyzed to remove any unreacted XG17K-RAFT that could be present in the samples. Dialysis was performed against deionized water using a dialysis membrane with a molar mass cut off of 50 kg mol−1 to ensure that only unreacted XG17K-RAFT and small molecules could cross the membrane. The first removal of water from the dialysis vessel was freeze-dried to assess how much, if any, XG17K-RAFT was removed from the latex samples. In all cases, the quantity was too low to be accurately determined (
