Article pubs.acs.org/Biomac

Temperature Responsive Cellulose-graf t-Copolymers via Cellulose Functionalization in an Ionic Liquid and RAFT Polymerization Andrea Hufendiek,†,‡,§ Vanessa Trouillet,∥ Michael A. R. Meier,*,§ and Christopher Barner-Kowollik*,†,‡ †

Preparative Macromolecular Chemistry, Institut für Technische Chemie und Polymerchemie, Karlsruhe Institute of Technology (KIT), Engesserstr. 18, 76128 Karlsruhe, Germany ‡ Institut für Biologische Grenzflächen and ∥Institute for Applied Materials (IAM) and Karlsruhe Nano Micro Facility (KNMF), Karlsruhe Institute of Technology (KIT), Herrmann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany § Laboratory of Applied Chemistry, Institute of Organic Chemistry, Karlsruhe Institute of Technology (KIT), Fritz-Haber-Weg 6, 76131 Karlsruhe, Germany S Supporting Information *

ABSTRACT: Well-defined cellulose-graf t-polyacrylamide copolymers were synthesized in a graf ting-f rom approach by reversible addition−fragmentation chain transfer polymerization (RAFT). A chlorine moiety (degree of substitution DS(Cl) ≈ 1.0) was introduced into the cellulose using 1-butyl3-methylimidazolium chloride (BMIMCl) as solvent before being substituted by a trithiocarbonate moiety resulting in cellulose macro-chain transfer agents (cellulose-CTA) with DS(RAFT) of 0.26 and 0.41. Poly(N,N-diethylacrylamide) (PDEAAm) and poly(N-isopropylacrylamide) (PNIPAM) were subsequently grafted from these cellulose-CTAs and the polymerization kinetics, the molecular weight characteristics and the product composition were studied by nuclear magnetic resonance spectroscopy, X-ray photoelectron spectroscopy, and size exclusion chromatography of the polyacrylamides after cleavage from the cellulose chains. The number-average molecular weights, Mn, of the cleaved polymers ranged from 1100 to 1600 g mol−1 for PDEAAm (dispersity Đ = 1.4−1.8) and from 1200 to 2600 g mol −1 for PNIPAM (Đ = 1.7−2.1). The LCST behavior of the cellulose-graft-copolymers was studied via the determination of cloud point temperatures, evidencing that the thermoresponsive properties of the hybrid materials could be finely tuned between 18 and 26 °C for PDEAAm and between 22 and 26 °C for PNIPAM side chains.

■

INTRODUCTION Cellulose is the most abundant natural organic polymer on earth (production of 1.5 × 1012 tons per year as part of biomass). It is a renewable, practically inexhaustibly available polymer that does not compete with food or feed.1 Although both cellulose and its derivatives have been used in various fields of application (i.e., clothing, packaging, building materials, foods, pharmaceuticals, separation techniques, cosmetics, sorption media) for years,1 it is highly desirable to identify novel cellulose derivatives to further increase the use of this polymer to substitute or complement polymeric materials based on crude oil. A promising approach toward new cellulose-based materials is homogeneous functionalization of cellulose in solution. So far, most (industrially relevant) cellulose derivatives are obtained from heterogeneous reactions, due to the insolubility of cellulose in most solvents. Homogeneous reactions, however, open new synthetic pathways for cellulosebased polymers with homogeneous substitutions patterns. Ideally, the homogeneous functionalization yields soluble products, while minimizing degradation of the cellulose backbone during functionalization. © 2014 American Chemical Society

Due to three hydroxyl groups per anhydroglucose unit (AGU), cellulose is most often functionalized by esterification or etherification. Since the discovery of ionic liquids (IL) as solvents for native cellulose,2 an increasing interest in these solvents (or IL/cosolvent mixtures) for cellulose functionalization can be noticed. In contrast to many other cellulose solvent systems (e.g., 4-methylmorpholine-N-oxide/water, dimethyl sulfoxide(DMSO)/tetrabutylammonium fluoride), ILs tolerate a vast variety of reactants, thus allowing for the synthesis of a wide variety of cellulose derivatives in solution.3,4 A detailed overview of cellulose derivatives synthesized in ILs was presented by Gericke et al. in 2012.5 Moreover, ILs are discussed as so-called “green” solvents due to their low vapor pressure. However, many aspects, such as detailed procedures for recycling and economical synthesis of these solvents, are still under investigation.5,6 Received: March 18, 2014 Revised: May 14, 2014 Published: May 15, 2014 2563

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

used as received. 1-Butyl-3-methylimidazolium chloride (BMIMCl; Aldrich, 95%) and Whatman filter paper No. 5 were stored in a desiccator over silica gel. 2,2′-Azobis(2-methylpropionitrile) (AIBN; Fluka, 98%) was recrystallized twice from ethanol. DEAAm (TCI, 98%) was passed over a short column of basic alumina and stored at −20 °C. NIPAM (TCI, 98%) was recrystallized twice from hexane and stored at −20 °C. Exemplary Synthesis of Cellulose-2-Chloropropionate (cellulose-Cl). A 2.5 wt % solution of cellulose in BMIMCl was prepared by cutting Whatman filter paper No. 5 into pieces of approximately 0.3 mm width, adding the pieces (0.80 g, 4.93 mmol anhydroglucose unit (AGU), 1.00 equiv) to the molten ionic liquid (31.2 g) and stirring overnight at 90 °C. The solution was cooled to 60 °C, and DBU (3.70 mL, 24.67 mmol, 5.00 equiv) in DMF (12.1 mL) was added dropwise. 2-Bromopropionyl bromide (2.10 mL, 19.74 mmol, 4.00 equiv) was dissolved in DMF (3.2 mL), and the resulting solution was slowly added dropwise to the reaction mixture while keeping the temperature between 10 and 15 °C by cooling with an ice/water bath. After complete addition, the reaction mixture was allowed to warm to ambient temperature and stirred for 24 h. The solution was poured into 50 vol % ethanol in deionized water (800 mL) resulting in precipitation of white floccules. The precipitate was filtered off and washed with deionized water (500 mL), ethanol (500 mL), and methanol (500 mL) and dried at 70 °C in vacuo to yield a slightly gray product (91−93% yield, with respect to cellulose). The DS of the 2chloropropionyl moiety (DS(Cl)) per AGU was determined by 1H NMR (Table 1). 1H NMR (400 MHz, DMSO-d6): δ (ppm) = 1.37− 1.90 (3H, CH3), 2.77−6.10 (8H, 7 × cellulose-H, CH-Cl). 13C NMR (100 MHz, DMSO-d6): δ (ppm) = 21.3 (CH3), 52.8 (CH-Cl), 64.3 (cellulose-C6), 71.7, 72.6, 74.0 (cellulose-C2, C3, C5), 79.1 (celluloseC4), 102.4 (cellulose-C1), 169.4 (O−CO).

Promising new cellulose derivatives are graf t-polymers, where polymer strands are attached to the cellulose backbone. In order to obtain such polymers, new (homogeneous) derivatization strategies for native cellulose have to be identified that allow the grafting of variable monomers from the cellulose backbone. Two general strategies exist: the so-called graf tingf rom approach,7−10 where polymer chains are grown directly from cellulose, and the grafting-to approach, where polymer strands are synthesized separately and subsequently ligated to the cellulose chains.11−14 Reversible deactivation radical polymerizations have already been successfully used for heterogeneous (surface) modification of cellulose in graf tingf rom15−25 and graf ting-to approaches26−28 and allow the attachment of polymer strands in a controlled fashion. Such materials are interesting for using cellulose to reinforce polymer matrices, where modification of the cellulose can increase the compatibility of the filler with the surrounding polymer. An overview of controlled polymerization techniques used for the preparation of such heterogeneously grafted cellulose was presented by Malmström and Carlmark in 2012.29 Moreover, (soluble) cellulose derivatives (e.g., hydroxyisopropylcellulose, ethyl hydroxylethyl cellulose), which are commercially available, have been employed for subsequent modification with variable polymer side chains (e.g., polystyrene, poly(vinyl acetate), poly(ethyl acrylate), PNIPAM, polyacrylamide) using RAFT polymerization.30−34 Ifuku et al. synthesized 2,3-di-O-methyl cellulose and subsequently grafted PNIPAM from this cellulose derivative using atom transfer radical polymerization (ATRP).35 Cellulose-graft-copolymers with poly(N,N-dimethylacrylamide) and polyacrylamide side chains have been prepared in LiCl/ N,N-dimethylacetamide (DMAc) by ATRP and single electron transfer living radical polymerization (SET-LRP). 36−38 Although a few reports on graf ting-from cellulose via ATRP in ILs or via functionalization of native cellulose in ILs exist,39−43 only one such report on graf ting-f rom via RAFT using native cellulose as starting material is available.44 In the current study, we present, to the best of our knowledge, the first solution based approach toward the synthesis of thermoresponsive cellulose-graf t-copolymers via RAFT polymerization using native cellulose as starting material. Thermoresponsive copolymers have received considerable attention during the last years and potential applications in, for example, drug delivery systems have been discussed.45,46 In such systems, the nontoxic, biocompatible, and biodegradable cellulose backbone may be of interest. In the current work, N,Ndiethylacrylamide (DEAAm) and N-isopropylacrylamide (NIPAM) have been used as monomers, leading to thermoresponsive cellulose-graf t-copolymers with a lower critical solution temperature (LCST). The kinetics, molecular weight characteristics, and reversible deactivation nature of the graft polymerization have been investigated in detail. LCST behavior was observed, and could be fine-tuned, in most products, depending on the molecular weight of the polyacrylamide side chains and characterized by determination of cloud point temperatures Tc.

■

Table 1. Comparison of the DS(Cl) of Different Synthesized Cellulose-Cls with Respect to the Applied Method for Determination of DS(Cl) via 1H NMR (see Characterization and Methods) cellulose-Cl

DS(Cl) method 1

DS(Cl) method 2

1 2 3 4

1.06 1.05 0.93 1.01

1.00 1.12 0.91 1.05

Exemplary Synthesis of Cellulose-Macro-RAFT Agent (Cellulose-CTA) with DS 0.26. Cellulose-Cl with a DS of 0.91 (0.80 g, 3.27 mmol cellulose-Cl repeating unit, 1.00 equiv) was dissolved in DMSO (20 mL) at 60 °C overnight. Methyl-3-mercaptopropionate (98.4 μL, 0.89 mmol, 0.27 equiv) and triethylamine (123.8 μL, 0.89 mmol, 0.27 equiv) were dissolved in DMSO (5.7 mL) at ambient temperature and stirred for 10 min. To this solution, carbon disulfide (144.4 μL, 2.40 mmol, 0.73 equiv) was added. After stirring for 10 min, the yellow solution was added to the cellulose-Cl solution at 40 °C and the reaction mixture was stirred overnight. Subsequently, the product was precipitated in deionized water (700 mL), filtered off and washed with water (350 mL), ethanol (450 mL), and methanol (450 mL) and dried at 70 °C in vacuo to afford a light yellow solid (for yields, see Table 3). For the synthesis of cellulose-CTA with a DS of 0.41 the cellulose-Cl was dissolved in DMSO (24 mL). Methyl-3-mercaptopropionate (166.8 μL, 1.51 mmol, 0.46 equiv) and triethylamine (209.8 μL, 1.51 mmol, 0.46 equiv) were dissolved in DMSO (5.8 mL) and after stirring for 10 min carbon disulfide (244.0 μL, 4.06 mmol, 1.24 equiv) was added. Otherwise, the procedure was the same as for the synthesis of cellulose-CTA with a DS of 0.26. The DS of the trithiocarbonate moiety (DS(RAFT)) was determined by 1H NMR analysis (Table 3). 1 H NMR (400 MHz, DMSO-d6): δ (ppm) = 1.37−1.90 (3H, CHCH3), 2.71−2.85 (2H, CH2−CO), 3.53−3.67 (5H, CH2−S, CH3− O), 2.71−6.10 (8H, 7 × cellulose-H, CH-Cl). 13C NMR (100 MHz, DMSO-d6): δ (ppm) = 16.3 (CH3-CH-S), (21.2 (CH3-CH-Cl), 31.6 (CH2-CH2-S), 32.0 (CH2-CH2-S), 47.5 (CH-S), 51.8 (O-CH3), 52.7

EXPERIMENTAL SECTION

Materials. 2-Bromopropionyl bromide (Alfa Aesar, 97%), carbon disulfide (Sigma-Aldrich, 99.9%), 1,8-diazabicyclo[5.4.0]-7-undecene (DBU; TCI, 98%), N,N-dimethylformamide (VWR Analpur), dimethyl sulfoxide (DMSO; Carl Roth, 99.8%), methyl-3-mercaptopropionate (Sigma-Aldrich, 98%), and triethylamine (Merck, 99%) were used as received. All other solvents were of analytical grade and 2564

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

(CH-Cl), 64.3 (cellulose-C6), 71.8, 72.6, 74.0 (cellulose-C2, C3, C5), 78.9 (cellulose-C4), 102.4 (cellulose-C1), 169.4 (OC-CH-Cl), 169.9 (OC-CH-S), 171.4 (OC-O-CH3), 221.5 (CS). Exemplary Procedure for Grafting of an Acrylamide from Cellulose-CTA. Cellulose-CTA with a DS of 0.26 (0.30 g, 0.27 mmol trithiocarbonate group, 1.00 equiv) was dissolved in DMF (38 mL) at 70 °C overnight. The solution was allowed to cool to ambient temperature and AIBN (9.0 mg, 0.05 mmol, 0.20 equiv) dissolved in DMF (4.2 mL) and DEAAm (1.37 g, 10.75 mmol, 39.42 equiv) or AIBN and NIPAM (1.60 g, 14.18 mmol, 51.95 equiv) dissolved in DMF (4.2 mL) were added. The reaction mixture was degassed with argon for 30 min and placed in an oil bath at 70 °C. Samples were removed from the polymerization mixture at preset time intervals, exposed to air, and cooled to −20 °C. Conversion was determined via 1 H NMR (see Characterization and Methods for details on evaluation). The polymer was collected by precipitation in cold diethyl ether. In the case of DEAAm, hexane was added to obtain a ratio of diethyl ether/hexane of 3:1. The precipitate was washed with diethyl ether (PNIPAM) or diethyl ether/hexane 3:1 (DEAAm) and dried at 70 °C in vacuo overnight. For removal of residual DMF and monomer, the samples were redissolved in DMSO and dialyzed against deionized water with a SpectraPor6 membrane (MWCO 3500) for 2 days at ambient temperature and lyophilized subsequently. Removal of homopolymer was investigated by SEC in DMF/LiBr (see Supporting Information). If necessary, the sample was redissolved in DMF, reprecipitated in diethyl ether/hexane 6:1 and dialyzed against water for 2 days (DEAAm) or redissolved in DMSO and precipitated in water at ambient temperature (NIPAM). In the case of NIPAM, water was gradually removed under reduced pressure at 55 °C until no further precipitation occurred. The cellulose-graf t-copolymers were obtained as light yellow or colorless solids. Procedure for hydrolysis of cellulose-graft-polyacrylamide copolymers. Cellulose-graft-polyacrylamide copolymers (15 mg) were dissolved in DMSO (0.7 mL) overnight. Aqueous sodium hydroxide (2 N, 8.2 mL) was added and the resulting suspension was stirred at 90 °C for 7 h and subsequently at 60 °C for 14 h. Hydrochloric acid (2 N) was added for neutralization and water was removed under reduced pressure at 55 °C. The residue was suspended in acetone (25 mL) overnight and the remaining solids were filtered off. After removal of acetone, the liquid residue was dialyzed against water with a SpectraPor6 membrane (MWCO 1000) for 24 h followed by lyophilization. Characterization and Methods. NMR measurements were performed on a Bruker Ascend spectrometer operating at 400 MHz for 1H nuclei and 100 MHz for 13C nuclei. Samples were stirred in DMSO-d6 at 60 °C until dissolution was observed (6−16 h). The chemical shifts, δ, were referenced to the solvent signal at 2.50 ppm. For determination of the conversion of DEAAm, the integrals corresponding to one vinylic proton (6.64−6.78 ppm) and the integral corresponding to the side chain methyl groups (0.70−1.34 ppm) were employed. The conversion of NIPAM was determined using the integral of two vinylic protons (5.97−6.26 ppm) and the integral of the hydrogen atoms of the side chain methyl groups (0.85− 1.32 ppm). Determination of DS(Cl) and DS(RAFT). Evaluation of DS(Cl) and DS(RAFT) was performed via 1H NMR analysis. The DS(Cl) was calculated using two methods. In the first case, the integral of the peak corresponding to the three protons a (Figure 3) of the methyl group (I(a)) and the integral of the peaks corresponding to the cellulose backbone protons 1−6 (I(AGU)) was used (method 1). In order to determine the value for I(AGU), the water peak was subtracted manually by integration of the respective peak.41,44

DS(Cl) =

7I(a) 3I(AGU)

the internal standard and protons a, the molar ratio of the 2chloropropionyl rest and the AGU could be calculated. Details regarding calculations for the second method can be found in the Supporting Information (method 2). An overview of the calculated values with respect to the applied method is given in Table 1. For DS(RAFT), the integral ratio of the resonance corresponding to the two protons f (I(f); Figure 3) and the resonance corresponding to the three protons a and a′ (I(a,a′)) was employed. In this case, DS(Cl)educt refers to the amount of 2-chloropropionyl groups per AGU in the corresponding cellulose-Cl educt.

DS(RAFT) =

3I(f) × DS(Cl)educt 2I(a , a′)

(2)

For the evaluation of the NMR spectra of cellulose-graf t-PDEAAm, benzoic acid was used as an internal standard for obtaining the ratio of polyacrylamide repeating units/AGU for one sample out of each set of samples. Subsequently, the change in the ratio of the integrals I(p)/ (I(i,a,a′) was employed for the calculation of respective values for the other samples of the set. The method is evaluated in the Results and Discussion section and details regarding the calculation are provided in the Supporting Information. NMR spectra of cellulose-graf t-PNIPAM were evaluated using the ratio of the integrals I(k)/I(AGU). Size Exclusion Chromatography (SEC). SEC with DMAc containing 0.03 wt % LiBr as eluent was performed for cleaved PDEAAm and PNIPAM side chains with a sample concentration of 2 g L−1 on a Polymer Laboratories PL-GPC 50 Plus Integrated System comprising an autosampler, a PLgel 5 μm bead-size guard column (50 × 7.5 mm), followed by three PLgel 5 μm Mixed-C columns (300 × 7.5 mm), and a refractive index detector at 50 °C with a flow rate of 1 mL min−1. The SEC system was calibrated against linear poly(methyl methacrylate) standards with molecular weights ranging from 700 to 2 × 106 Da. The samples were filtered through polytetrafluorethylene (PTFE) membranes with a pore size of 0.2 μm prior to injection. SEC of the cellulose-graf t-copolymers was performed at a sample concentration of 1 g L−1 with DMF containing 1 g L−1 LiBr as eluent at 50 °C on an Agilent Series 1200 HPLC system including a isocratic pump (G1310A), an autosampler (G1329A), a thermostat controlled column compartment (G1316A) and a refractive index detector (G1362A) at a flow rate of 0.8 mL min−1. Separation was achieved on a SEC column (PSS, GRAM analytical, 300 × 8.00 mm, 10 μm particle size, 3000 Å porosity) with precolumn (50 × 8.00 mm). Linear poly(methyl methacrylate) standards with molecular weights ranging from 4300 to 3.73 × 106 g mol−1 were used for calibration. The samples were dissolved in the eluent at 60 °C for 14 h and filtered through PTFE membranes with a pore size of 0.45 μm prior to injection. SEC data of Whatman filter paper No. 5, cellulose-Cl, and cellulose-CTA was obtained using the same SEC system as for cellulose-graft-copolymers with DMAc/LiCl (10 g L−1 LiCl) as eluent at a flow rate of 0.5 mL min−1 and a sample concentration of 1 g L−1. The column compartment was heated to 70 °C, and the measurement cell of the refractive index detector was kept at 50 °C.48,49 Calibration was performed with the same linear poly(methyl methacrylate) standard as for cellulose-graft-copolymers. Dissolution of Whatman filter paper No. 5 was achieved in a solvent exchange procedure from literature (refer to Supporting Information for detailed information).48,49 Cloud Point Determination. Cloud points were measured on a Cary 300 Bio UV/vis spectrophotometer (Varian) at 500 nm in the temperature range from 7 to 60 °C at a heating rate of 0.5 °C min−1. For sample preparation, a solution of polymer in DMSO with a concentration of 10 g L−1 was prepared by stirring overnight at 60 °C. A total of 100 μL of this solution were added to 3.9 mL of water while stirring vigorously and cooling with an ice bath, resulting in an aqueous solution with a sample concentration of 0.25 g L−1. X-ray Photoelectron Spectroscopy (XPS). XPS investigations were performed on a K-alpha spectrometer (ThermoFisher Scientific, East Grinstead, U.K.) using a microfocused, monochromated Al Kα Xray source (400 μm spot size). The kinetic energy of the electrons was measured by a 180° hemispherical energy analyzer operated in the

(1)

A second method for the determination of DS(Cl) was adopted from the literature.47 Here, a defined amount of benzoic acid was added as an internal standard to a defined amount of cellulose-Cl sample and dissolved in DMSO. Using the integral ratio of the aromatic protons of 2565

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

constant analyzer energy mode (CAE) at 50 eV pass energy for elemental spectra. The photoelectrons were detected at an emission angle of 0° with respect to the normal of the sample surface. The Kalpha charge compensation system was employed during analysis, using electrons of 8 eV energy and low-energy argon ions to prevent any localized charge build-up. Data acquisition and processing using the Thermo Avantage software is described elsewhere.50 The spectra were fitted with one or more Voigt profiles (BE uncertainty: ±0.2 eV). The analyzer transmission function, Scofield51 sensitivity factors, and effective attenuation lengths (EALs) for photoelectrons were applied for quantification. EALs were calculated using the standard TPP-2 M formalism.52 All spectra were referenced to the C 1s peak (C−C, C− H) at 285.0 eV binding energy controlled by means of the well-known photoelectron peaks of metallic Cu, Ag, and Au, respectively. The powdery samples were placed onto a sticky carbon pad as a closed layer. In the case of Whatman filter paper No. 5, a square piece of approximately 10 mm width was fixed on the sample stage and then analyzed.

of the cellulose backbone, but the determined molecular weights should not be discussed as absolute values, since the calibration standards are of a very different nature than the investigated samples. SEC data of native cellulose and celluloseCl in DMAc/LiCl as eluent does not show a reduction in peak molecular weight (Figure 2).

■

RESULTS AND DISCUSSION Synthesis of Cellulose-CTAs. In the first step toward the synthesis of cellulose-CTAs, an α-halogen cellulose-ester was synthesized in the IL BMIMCl (Figure 1) in a procedure adopted from the literature.39−41 Figure 2. SEC data for Whatman filter paper no. 5 (solid line, Mn = 40800 g mol−1, Mp = 88900 g mol−1, Đ = 1.93), cellulose-Cl 1 (dashed line, Mn = 82800 g mol−1, Mp = 154900 g mol−1, Đ = 5.41), and cellulose-CTA 1 (dotted line, Mn = 59800 g mol−1, Mp = 87200 g mol−1, Đ = 4.69). SEC was performed in DMAc/LiCl (10 g L−1 LiCl) at 70 °C relative to PMMA standards.

The determined molecular weight of the cellulose-Cl is significantly higher than the molecular weight of native cellulose, which may be due to functionalization but also to a different relationship between the hydrodynamic volume and the degree of polymerization of functionalized cellulose in the eluent. The data verifies that the presented cellulose esterification is mild and does not result in pronounced degradation of the cellulose backbone. To investigate the formation of an α-halogen cellulose-ester by an atom sensitive characterization method, XPS measurements were performed. XPS is a powerful tool for surface analysis and has been used in our group for both the characterization of cellulose surfaces and for surface bound trithiocarbonates.28,53,54 The general trends, which were already observed by NMR spectroscopy for the synthesis of cellulose-Cl, are underpinned by the XPS analysis. No chlorine or bromine is detected in Whatman filter paper No. 5, but for cellulose-Cl 1 and cellulose-Cl 2 (Table 2), similar chlorine contents of 5.6 atomic% and 5.8 atomic%, respectively, are observed (Figure S9). Yet, the results (Table 2) clearly show that in contrast to the reports in the literature,39−43 which do not present data from atom sensitive characterization methods (such as XPS or elemental analysis), no bromine moieties can be detected in the cellulose-Cl samples. This can be explained by a complete replacement of the bromide moieties by chlorine anions of the solvent in a nucleophilic substitution reaction. It is highly likely that the replacement also takes place if AMIMCl as solvent is used. Although this replacement does not seem to interfere with common consecutive reactions, such as nucleophilic substitution reactions or ATRP polymerization, an atomsensitive characterization method yields valuable additional information, and nucleophilic replacements by solvent anions

Figure 1. Synthetic pathway to cellulose-graf t-polyacrylamide copolymers via RAFT-polymerization.

Table 1 collates the DS(Cl) data of the synthesized celluloseCls used in the current work. Reproducible values, that is, 0.91 < DS(Cl) < 1.12, were obtained by both methods for the determination of DS(Cl) (see Characterization and Methods), and the two methods can therefore be considered equivalent. Furthermore, the values of DS(Cl) obtained are in good agreement with the literature data for the functionalization of cellulose with 2-bromopropionyl bromide or 2-bromo isobutyryl bromide in the IL 1-allyl-3-methylimidazolium chloride (AMIMCl).39−43 For a direct comparison of native cellulose and its derivatives, SEC in DMAc/LiCl, calibrated against PMMA standards, was employed. The method proved to be very useful to investigate the influence of the functionalization procedure on degradation 2566

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

Table 2. NMR and XPS Data for Cellulose-Cl, Cellulose-CTA, and the Native Cellulose Starting Materiala DS(Cl)NMR filter paper cellulose-Cl 1 cellulose-Cl 2 cellulose-CTA 1 cellulose-CTA 2

1.00 1.12 (0.73)b (0.71)b

DS(RAFT)NMR

0.27 0.41

(S/Cl)NMR

ClXPS (atomic%)

SXPS (atomic%)

(S/Cl)XPS

1.1 1.7

0 5.6 5.8 4.4 4.2

0 0 0 2.6 4.5

0 0 0 0.6 1.1

a

The corresponding spectra are given in the Supporting Information. bValues are calculated by subtraction of DS(RAFT)NMR from DS(Cl)NMR of cellulose-Cls.

should always be taken into account when synthesizing cellulose derivatives in ionic liquids.55 Moreover, follow-up reactions can be planned more precisely depending on the reactive moiety (in this case, Cl vs Br, which gives, for instance, a different initiation behavior in ATRP). In the second step, the chlorine moiety of cellulose-Cl was partially replaced by trithiocarbonate functional groups in a nucleophilic substitution reaction in DMSO using a procedure adopted from the literature.56 The choice of DS(RAFT) is mainly limited by the solubility of the resulting cellulose-CTAs. Test reactions showed that a DS(RAFT) of more than approximately 0.46 leads to insoluble products. Thus, cellulose-CTAs with a DS(RAFT) of 0.26 and 0.41 (determined by 1H NMR) were synthesized (refer to Figure 3, Table 3). New proton resonances, corresponding to methyl

Table 3. Obtained DS(RAFT) Values for Synthesized Cellulose-CTAs Determined by 1H NMR (see Characterization and Methods)a celluloseCTA 1 2 3 4

DS(Cl) (cellulose-Cl) 1.00 (celluloseCl 1) 1.12 (celluloseCl 2) 0.91 (celluloseCl 3) 1.05 (celluloseCl 4)

methyl-3mercaptopropionate (equiv/AGU)

DS(RAFT)

yieldb (%)

0.27

0.27

85.6

0.46

0.41

73.2

0.27

0.26

84.1

0.46

0.41

81.9

a

DS(Cl) is given for the cellulose-Cl from which the corresponding cellulose-CTA is synthesized. bThe yield is given with respect to cellulose.

introduction of trithiocarbonate groups is also confirmed by C NMR spectroscopy (Figure 4).

13

Figure 3. 1H NMR spectra of cellulose-CTA 4 (top) with a DS(RAFT) of 0.41 and the corresponding cellulose-Cl 4 with a DS(Cl) of 1.05 in DMSO-d6. The assignment of the proton resonances is depicted in Figure 1.

protons h and methylene protons e and f, which are absent in spectra of cellulose-Cl, are found for cellulose-CTAs, verifying the successful attachment of the trithiocarbonate groups on the cellulose backbone. The resonance corresponding to methylene protons f, which is separated from the cellulose backbone resonances, was used for the calculation of DS(RAFT). The resonance corresponding to methyl protons a′ overlaps with the signal of protons a, resulting in a broadening of the peak corresponding to these protons (Figure 3). The overview of cellulose-CTAs given in Table 3 evidences the quantitative introduction of the employed amount of thiol into the cellulose backbone for a DS(RAFT) of 0.26, whereas 90% of the employed thiol is present in the cellulose-CTA with a higher DS(RAFT) of 0.41. This may be due to increased steric hindrance. The successful

Figure 4. 13C NMR spectra of cellulose-CTA 4 (top) with a DS(RAFT) of 0.41 and the respective cellulose-Cl 4 with a DS(Cl) of 1.05 in DMSO-d6. Annotation of carbon atoms is given in Figure 1.

Especially the resonance corresponding to the carbon atom of the trithiocarbonate group d confirms a successful synthesis of cellulose-CTA, since it can be assigned uniquely to trithiocarbonates.57 Moreover, the resonances corresponding to the methine carbon and the methyl carbon of the 2chloropropionyl moiety (b, a), as well as the CTA group (b′, a′), can be distinguished in the 13C NMR spectra, thus, clearly confirming the partial substitution. The change in chemical shift is less pronounced for carbonyl carbon atoms c and c′. Furthermore, the additional resonances of the methylene 2567

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

carbon atoms e and f, carbonyl carbon atom g, and methyl carbon atom h can be assigned in the cellulose-CTA spectrum. The two cellulose-CTAs, cellulose-CTA 1 and cellulose-CTA 2, were additionally investigated by XPS (refer to Table 2). In both filter paper and cellulose-Cls, no sulfur was detected, whereas the S 2p signal of cellulose-CTA clearly shows two doublets (Figure S10). It is known from the literature that the doublet associated with S 2p3/2 at 162.6 eV (CS) and the second doublet corresponding to S 2p3/2 at 164.0 eV (C−S) can be attributed to trithiocarbonate moieties.58 Cellulose-CTA 2 shows a 1.7 times higher sulfur content than cellulose-CTA 1, confirming the successful introduction of trithiocarbonate moieties and the general trends when comparing the samples, that is, a higher amount of trithiocarbonate moieties present in cellulose-CTA 2. In addition, the C 1s signal (Figure S8) clearly demonstrates the synthesis of cellulose-Cl and cellulose-CTA, since the intensity of the C 1s components at 285.0 eV (C−C, C−H) and at 289.2 eV (OC−O) increased in comparison to the peak at 286.7 eV (C−O), which characterizes the cellulose.26 SEC of cellulose-CTA shows a reduction in molecular weight as compared to cellulose-Cl (Figure 2). Nevertheless, since no reduction in peak molecular weight was observed in comparison with native cellulose, degradation of the cellulose backbone during the synthesis of cellulose CTA is very likely negligible. Grafting of Polyacrylamides from Cellulose-CTA. The acrylamide monomers NIPAM and DEAAm were grafted from these cellulose-CTAs in DMF at 70 °C by RAFT polymerization (Figure 1). Although the cellulose-CTAs are better soluble in DMSO, DMF was chosen as a solvent for polymerization since it yielded better and more reproducible results in first test reactions. Cellulose-CTAs are soluble (see, for example, Figure S1), but small, swollen, gel-like particles remain. However, these particles can be well dispersed in the reaction mixture. In total, four cellulose-graft-polyacrylamide copolymers were investigated. For the two cellulose-CTAs with different DS(RAFT) of 0.26 and 0.41, both a cellulose-graf tPNIPAM and a cellulose-graf t-PDEAAm copolymer were prepared. An overview of the used ratio of reactants and the monomer concentration is given in Table 4. The kinetic plot for the four polymerizations is shown in Figure 5. Samples for the determination of the conversion by 1H NMR were withdrawn every hour within a time span of 7 h. All four graf t-copolymers show the pseudo-first-order kinetics expected for RAFT polymerization during the first 4 h of polymerization. However, for all samples, the conversion drastically levels off after 4 h of polymerization. After 6−7 h, a plateau value is reached and the polymerization ceases. There

Figure 5. Kinetic plot for the graf ting-f rom RAFT polymerization of DEAAm and NIPAM using a cellulose-CTA with a higher DS(RAFT) of 0.41 and a cellulose-CTA with a lower DS(RAFT) of 0.26.

are several effects that have to be taken into account, resulting in such a significant reduction in the rate of polymerization and eventually stopping the polymerization completely. First of all, the cellulose backbone would be crowded with polyacrylamide coils after a certain time, leading to increased steric hindrance and reduced availability of RAFT groups, an effect that can contribute to a reduction in polymerization rate (reduced reinitiation). Furthermore, a reduction in the rate of polymerization can occur due to a reduction in radical concentration by termination. Two effects that result in increased termination are plausible for the presented graf t-copolymers. The CTA-groups are in very close proximity with in average less than three (DS(RAFT) = 0.41) or less than four (DS(RAFT) = 0.26) AGUs without CTA groups as spacers between them, resulting in close proximity of radicals, thus, facilitating termination and resulting in ring formation.59 Coupling of two R-groups of the RAFT agent (cross-linking) is also possible.27 For PDEAAm side chains, the conversion levels off at 70% for DS(RAFT) = 0.41 and at 80% for DS(RAFT) = 0.27. In general, higher conversions and faster polymerization are observed for DEAAm. For NIPAM, the conversion levels off at 50% (DS(RAFT) = 0.26) or at 65% (DS(RAFT) = 0.41). The difference in rate between the two polymer systems may be attributed to the different nature of the monomers (rates of propagation), yet the effect of DS(RAFT) on the conversion is less clear and the tendency is inverted for the two monomers. It should also be mentioned that the experimental error in the system is relatively high, mainly due to the state of dissolution of the cellulose-CTA, which might be influenced by the DS(RAFT). To obtain information on the attachment of polyacrylamide side chains, 1H NMR spectroscopy of the cellulose-graf tcopolymers was performed. Exemplary spectra are given in the Supporting Information (refer to Figure S5), confirming the successful grafting of polyacrylamide chains from the cellulose backbone. Resonances corresponding to the polyacrylamide backbone (i,j) are present for both PDEAAm and PNIPAM. Polyacrylamide side group methyl protons p and l can be assigned for PDEAAm and PNIPAM, respectively. In the case of PDEAAm side chains, the resonance of the four methylene protons o overlaps with the resonances of cellulose, for PNIPAM the methine proton resonance m overlaps with cellulose resonances. For PNIPAM side chains the amide proton resonance at 7.5 ppm is clearly visible and separated

Table 4. Overview over Monomer Concentrations and Reactant Ratios for the Synthesized Cellulose-graf tPolyacrylamide Polymers cellulose-graf t-copolymer cellulose-graf t-PDEAAm, DS(RAFT) = 0.27 (P1) cellulose-graf t-PDEAAm, DS(RAFT) = 0.41 (P2) cellulose-graf t-PNIPAM, DS(RAFT) = 0.26 (P3) cellulose-graf t-PNIPAM, DS(RAFT) = 0.41 (P4)

[M]0 (mol L−1)

[M]0/[CTA]/[AIBN]

0.25

39/1/0.2

0.24

40/1/0.2

0.34

52/1/0.2

0.34

55/1/0.2

2568

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

from other resonances. Evaluation of NMR data with respect to the molecular weight of the polyacrylamide side chains will be discussed in detail below. To investigate the relationship between molecular mass and conversion, the polyacrylamide side chains were cleaved from the cellulose backbone after purification (refer to Figure S2). Prior to hydrolysis, complete removal of homopolymer was checked by SEC (refer to Figure S1). For P4, separation of the homopolymer was not possible due to the similar solubility of homopolymer and graf t-copolymer, resulting in either very significant loss of graf t-copolymer or incomplete removal of homopolymer. These graf t-copolymers are therefore excluded from further discussion. SEC traces of the hydrolyzed side chains of P3 are given in Figure 6. Figure 7. Evolution of Mn (full symbols), Mp (open symbols), as well as Đ (crossed symbols) of polyacrylamide side chains with conversion. Triangular symbols refer to P3. The other symbols refer to PDEAAm side chains (P1 (circles), P2 (squares)). The dashed (P1), solid (P2), and dotted (P3) lines refer to theoretical values for Mn. Refer to the Supporting Information section for the calculation of the theoretical values.

time, leading to the observed high values for dispersity for both NIPAM and DEAAm. The evolution of Mp as a function of conversion clearly follows a linear trend, indicating that a significant amount of side chains actually grows in a controlled fashion. For PNIPAM side chains and DS(RAFT) = 0.26, Mn increases linearly with conversion and the deviation from the calculated molecular weights, Mtheo, is small. It is common practice to neglect the formation of polymer chains derived from initiator fragments (homopolymer formation) when calculating the theoretical molecular weights (see Supporting Information) of RAFT polymers.60 In the case of the PNIPAM side chains, the calculation corresponds well with the experimental findings. For PDEAAm side chains, the experimentally determined molecular weights are smaller than Mtheo. Possibly, this may be due to an increased formation of homopolymer in the case of PDEAAm. In general, determination of Mn from NMR data is more difficult and less reproducible than evaluation of SEC data of cleaved polyacrylamide side chains for the presented polymers, which is due to the broadness and overlap of resonances as well as limited sensitivity of NMR. Since the signals corresponding to proton resonances p and i, which are used for evaluation of PDEAAm copolymers using an additional standard (see Characterization and Methods and Supporting Information), are less separated in PNIPAM copolymers (Figure S5), the resonance of proton k and of the cellulose backbone are used for evaluation of PNIPAM copolymers. As shown above for cellulose-Cls, these methods can be considered to be equivalent (Table 1). In general, the amount of polyacrylamide repeating units per AGU was determined first (see Supporting Information, Figure S6). For a better comparison with SEC data, these values were divided by DS(RAFT), thus, assuming that all RAFT groups participate in the polymerization (Figure 8). For PDEAAm side chains, values for Mn,NMR are generally higher than the values determined by SEC, attributed to general difficulties with the NMR analysis of polymers (see above), which yields absolute values with large errors for PDEAAm where no resonance is as clearly separated as in PNIPAMcopolymers (proton k). As expected, the error increases with increasing number of polymer repeating units, leading to the

Figure 6. SEC traces of PNIPAM side chains after hydrolysis of the cellulose-graft-PNIPAM copolymer with DS(RAFT) = 0.26 (P3). SEC traces are given with respect to the time of polymerization.

Similar plots for PDEAAm side chains are included in the Supporting Information (see Figures S3 and S4). The increase in molecular weight of the side chains with increase in polymerization time and conversion is confirmed for all presented polymers. Linear evolution of the number molecular weight, Mn, as a function of conversion is an important characteristic of reversible deactivation polymerizations. For the hydrolyzed polyacrylamide side chains, the evolution of molecular weight and dispersity as a function of conversion is displayed in Figure 7. In contrast to what is expected for RAFT polymerizations (Đ ≈ 1.0−1.2), the dispersity Đ does not decrease with conversion. For PDEAAm side chains, Đ levels off at a value of 1.7. For PNIPAM side chains, a continuous increase in Đ up to a value greater than 2.0 was observed. The investigation of the molecular weight of PDEAAm side chains shows the same tendency for the evolution of molecular weight as a function of conversion for both cellulose-graft-copolymers, independent of DS(RAFT), indicating that the kinetics of the polymerization mainly depend on the choice of monomer and CTA, whereas the influence of DS(RAFT) is significantly smaller. The calculated molecular weights are much higher than the experimentally determined values for Mn. Moreover, Mn levels off after 3 to 4 h of polymerization and reaches a plateau value, possibly associated with the leveling off of conversion, but also with a continuous entering of new RAFT groups into the polymerization. Due to the aforementioned steric hindrance and limited mobility of the CTA end groups on the cellulosebackbone, they may not all start the polymerization at the same 2569

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

Figure 9. SEC traces of polyacrylamide side chains of cellulose-graftPNIPAM starting material P3 before (50.8% conversion, 3 h) and after extension of the side chains (dotted line, 47.3% conversion, 3 h) with PNIPAM and block copolymer formation with DEAAm (dashed line, 55.9% conversion, 3 h).

Figure 8. Comparison of Mn values obtained by 1H NMR spectroscopy (open symbols) and SEC (full symbols). Triangular symbols refer to P3, other symbols refer P1 (circles) and P2 (squares). Theoretical values are deduced in the same fashion as in Figure 7.

increased deviation of the three last values at highest conversion for P2 (Figure 8). These values will thus be excluded from discussion. Despite a deviation in absolute values, the NMR data confirms the trends already observed via SEC for PDEAAm side chains, that is, initial linear increase of Mn with conversion, but slight leveling off at higher conversions. Moreover, the NMR data clearly shows that a larger DS(RAFT) leads to more DEAAm units on the cellulose backbone (refer to Figure S6). Taking the different DS(RAFT) into account leads to calculated Mn,NMR values, which fall onto the same trend line. The difference between calculated and experimentally determined values of molecular weights of PDEAAm side chains was already discussed above for the SEC data. XPS was performed for selected cellulose-graft-PDEAAm copolymers. While the XPS data clearly confirms the attachment of polyacrylamide side chains by detection of nitrogen (main peak N 1s at 399.9 eV corresponding to amide and amine functional groups,61 Figure S11), no trend for the nitrogen content as a function of polymerization time or conversion can be detected (refer to Table S1), which is most likely due to the relatively low increases in Mn of the polyacrylamide side chains. All measured cellulose-graf t-polyacrylamide copolymers feature a nitrogen content of 5−7 atomic%. An important aspect of reversible deactivation polymerizations is that reinitiation of the chain growth and block copolymer formation must be possible. Chain elongation with NIPAM and block copolymer formation with DEAAm was therefore investigated for P3 using the same initial monomer concentrations as for the starting material formation. 1H NMR spectroscopy of the cellulose-graf t-polyacrylamide copolymer after block copolymer formation (Figure S7) confirms the presence of PDEAAm after this experiment. The SEC traces of the hydrolyzed side chains also clearly confirm successful reinitiation of the chain growth for both chain elongation and block copolymer formation (Figure 9). However, it is likely that not all side chains are reinitiated, since a tailing of the SEC traces at low molecular weights for the reinitiated chains and an increase in dispersity was observed, which is probably due to the aforementioned steric hindrance. Interestingly, the trends detected for the graf tingf rom polymerization of NIPAM and DEAAm are also observed for chain elongation and block copolymer formation. For chain

elongation, the Mp is doubled, as expected for the same initial monomer concentrations and similar conversion (50.8 and 47.3%) in both graf ting-f rom polymerization and chain elongation. Before chain elongation, the Mp of the PNIPAM side chains reads 5800 g mol−1 and after chain elongation the Mp is 10600 g mol−1, which corresponds very well to a controlled character of the polymerization. In contrast, the conversion for DEAAm in block copolymer formation is higher (55.9%) at the same polymerization time (3 h), but the resulting Mp of the block copolymer is lower than for the elongated homopolymer (7300 g mol−1). The collated evidence clearly indicates that a rather controlled RAFT polymerization took place, which is remarkable considering the complexity and the overall high molecular weight of the studied system. Investigation of the Thermoresponsive Behavior of Cellulose-graf t-Polyacrylamides. The thermoresponsive behavior of cellulose-graf t-polyacrylamide polymers in water was subsequently investigated. The cellulose-graf t-copolymers are soluble in water at low concentrations (≈0.25 g L−1). Due to a coil to globule transformation, a solution of a polymer with LCST behavior becomes turbid at a certain critical temperature, the cloud-point temperature Tc (refer to Figure S12). This temperature was determined as the point of inflection of transmission measurements at a fixed wavelength using a UV/ vis spectrometer. Turbidity measurements for P2 polymers are presented in Figure 10. For turbidity data of P1 and P3, refer to Figures S13 and S14 in the Supporting Information. The obtained data sets show a reduction in Tc with decreasing molecular weight and lower values for Tc in cooling cycles.45,46 For P2, the Tc increases from 15 °C/18 °C (cooling/heating) for Mp = 2100 g mol−1 to 21 °C/25 °C for Mp = 3100 g mol−1. Although it is known from the literature that for homopolymers a decrease in LCST is commonly observed with increase in molecular weight,45 an increase in Tc values with increasing molecular weight corresponds to earlier results from our research group for PDEAAm with hydrophobic end groups.62 The Tc of cellulose-graf t-copolymer presented in the current work is very sensitive to changes in molecular weight of side chains, since an increase in Mp of only eight repeating units results in an increase in Tc of 6−7 °C. However, the transition is not as sharp as for linear homopolymers,62 which can be attributed to the influence of the molecular weight of the 2570

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

are also better soluble in water at lower molecular weights of the side chains compared to PNIPAM copolymers.

■

CONCLUSIONS To the best of our knowledge, we present the first approach toward thermoresponsive cellulose-graf t-polyacrylamide copolymers via the functionalization of cellulose in an ionic liquid and subsequent RAFT polymerization. Cellulose-Cl and cellulose-CTAs with two DS(RAFT) were synthesized without significant degeneration of the cellulose backbone and characterized by 1H NMR spectroscopy, 13C NMR spectroscopy, SEC, and XPS analysis. Two polyacrylamides, PDEAAm and PNIPAM, were grafted from the cellulose backbone. Furthermore, the kinetics and reversible deactivation nature of the polymerization were studied by 1H NMR and SEC. The polymerizations level off at a certain limiting value of the conversion. Yet, a linear increase of molecular weight with increasing conversion, reinitiation of the polymerization and block copolymer formation was shown for the cellulose-graf tcopolymers. The synthesized cellulose-copolymers are thermoresponsive and show LCST behavior in water, which is very sensitive toward the molecular weight of the side chains. The cellulose-graft-copolymers were soluble in water already at low molecular weights of the side chains. The presented solutionbased synthesis of cellulose-graf t-polyacrylamides, employing RAFT polymerization, takes place in a controlled fashion and may be transferred to the synthesis of a variety of other cellulose-graft-copolymers. The current study thus provides a benign and efficient access route to cellulose hybrid materials, introducing a synthetic platform technology that may be exploited to employ cellulose materials in applications where a stimuli-response is of importance (e.g., delivery system).

Figure 10. Turbidity measurements for P2 at a heating/cooling rate of 0.5 °C min−1 for different polymerization times. Solid lines show turbidity measurements while cooling, whereas dotted lines refer to data obtained during heating.

polyacrylamide chains on the LCST as well. In the present systems not only the polyacrylamide side chains show a relatively high dispersity yet also the cellulose backbone (Đ = 1.93), resulting in broad transmission curves. The data obtained from turbidity measurements is summarized for cellulose-graf tpolyacrylamides in Figure 11 and in Table S2. Interestingly, for

■

ASSOCIATED CONTENT

S Supporting Information *

Additional SEC, NMR, and XPS data as well as additional information on thermoresponsive behavior. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]; m.a.r.meier@kit. edu. Web: www.macroarc.de; www.meier-michael.com.

Figure 11. Cloud point temperatures (Tc) determined by UV/vis measurements for cellulose-graf t-polyacrylamide copolymers with different side chains and different DS(RAFT). Values for Tc are given for heating cycles (Tc↑). Solid lines are included to guide the eye.

Notes

The authors declare no competing financial interest.

■

PDEAAm side chains, the Tc only depends on the molecular weight of the polyacrylamide side chains and not on DS(RAFT), which is in good agreement with afore discussed SEC and NMR data. Most likely, the difference in DS(RAFT) for the presented cellulose-graf t-copolymers is too small to result in a significantly different thermal behavior of the polymers. At 1700 g mol−1 < Mp < 2100 g mol−1 the PDEAAm-copolymers become insoluble. Within a change in Mp from 2100 to 3500 g mol−1 (≈12 repeating units), the Tc increases about 8 °C. In contrast, PNIPAM copolymers become insoluble at 3500 g mol−1 < Mp < 5200 g mol−1 and an increase from 5200 to 7100 g mol−1 (≈17 repeating units) in Mp leads to an increase of 4− 5 °C in Tc. Therefore, the thermoresponsive behavior of cellulose-graf t-PDEAAm copolymers is not only more sensitive toward the molecular weight of side chains, but these polymers

ACKNOWLEDGMENTS A.H. thanks the Fond der Chemischen Industrie for a scholarship funding her Ph.D. studies. C.B.-K. acknowledges continued support from the Karlsruhe Institute of Technology (KIT) via its Helmholtz BioInterfaces program.

■

REFERENCES

(1) Klemm, D.; Heublein, B.; Fink, H.-P.; Bohn, A. Angew. Chem., Int. Ed. 2005, 44 (22), 3358−3393. (2) Swatloski, R. P.; Spear, S. K.; Holbrey, J. D.; Rogers, R. D. J. Am. Chem. Soc. 2002, 124 (18), 4974−4975. (3) El Seoud, O. A.; Koschella, A.; Fidale, L. C.; Dorn, S.; Heinze, T. Biomacromolecules 2007, 8 (9), 2629−2647. (4) Seoud, O.; Heinze, T. Organic Esters of Cellulose: New Perspectives for Old Polymers. In Polysaccharides I; Heinze, T., Ed.; Springer: Berlin Heidelberg, 2005; Vol. 186, pp 103−149.

2571

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572

Biomacromolecules

Article

(5) Gericke, M.; Fardim, P.; Heinze, T. Molecules 2012, 17 (6), 7458−7502. (6) Zhu, S.; Wu, Y.; Chen, Q.; Yu, Z.; Wang, C.; Jin, S.; Ding, Y.; Wu, G. Green Chem. 2006, 8 (4), 325−327. (7) Carlsson, L.; Malmstrom, E.; Carlmark, A. Polym. Chem. 2012, 3 (3), 727−733. (8) Habibi, Y.; Goffin, A.-L.; Schiltz, N.; Duquesne, E.; Dubois, P.; Dufresne, A. J. Mater. Chem. 2008, 18 (41), 5002−5010. (9) Labet, M.; Thielemans, W. Polym. Chem. 2012, 3 (3), 679−684. (10) Yang, Q.; Pan, X.; Huang, F.; Li, K. Cellulose 2011, 18 (6), 1611−1621. (11) Månsson, P.; Westfelt, L. J. Polym. Sci., Polym. Chem. Ed. 1981, 19 (6), 1509−1515. (12) Biermann, C. J.; Chung, J. B.; Narayan, R. Macromolecules 1987, 20 (5), 954−957. (13) Tsubokawa, N.; Iida, T.; Takayama, T. J. Appl. Polym. Sci. 2000, 75 (4), 515−522. (14) Roy, D.; Semsarilar, M.; Guthrie, J. T.; Perrier, S. Chem. Soc. Rev. 2009, 38 (7), 2046−2064. (15) Carlmark, A.; Malmström, E. J. Am. Chem. Soc. 2002, 124 (6), 900−901. (16) Carlmark, A.; Malmström, E. E. Biomacromolecules 2003, 4 (6), 1740−1745. (17) Perrier, S.; Takolpuckdee, P.; Westwood, J.; Lewis, D. M. Macromolecules 2004, 37 (8), 2709−2717. (18) Lee, S. B.; Koepsel, R. R.; Morley, S. W.; Matyjaszewski, K.; Sun, Y.; Russell, A. J. Biomacromolecules 2004, 5 (3), 877−882. (19) Roy, D.; Guthrie, J. T.; Perrier, S. Macromolecules 2005, 38 (25), 10363−10372. (20) Roy, D. Aust. J. Chem. 2006, 59 (3), 229−229. (21) Barsbay, M.; Güven, O.; Stenzel, M. H.; Davis, T. P.; BarnerKowollik, C.; Barner, L. Macromolecules 2007, 40 (20), 7140−7147. (22) Roy, D.; Knapp, J. S.; Guthrie, J. T.; Perrier, S. Biomacromolecules 2008, 9 (1), 91−99. (23) Lindqvist, J.; Nyström, D.; Ö stmark, E.; Antoni, P.; Carlmark, A.; Johansson, M.; Hult, A.; Malmström, E. Biomacromolecules 2008, 9 (8), 2139−2145. (24) Hansson, S.; Ö stmark, E.; Carlmark, A.; Malmström, E. ACS Appl. Mater. Interfaces 2009, 1 (11), 2651−2659. (25) Zampano, G.; Bertoldo, M.; Bronco, S. Carbohydr. Polym. 2009, 75 (1), 22−31. (26) Hansson, S.; Trouillet, V.; Tischer, T.; Goldmann, A. S.; Carlmark, A.; Barner-Kowollik, C.; Malmström, E. Biomacromolecules 2012, 14 (1), 64−74. (27) Barner, L.; Davis, T. P.; Stenzel, M. H.; Barner-Kowollik, C. Macromol. Rapid Commun. 2007, 28 (5), 539−559. (28) Goldmann, A. S.; Tischer, T.; Barner, L.; Bruns, M.; BarnerKowollik, C. Biomacromolecules 2011, 12 (4), 1137−1145. (29) Malmström, E.; Carlmark, A. Polym. Chem. 2012, 3 (7), 1702− 1713. (30) Stenzel, M. H.; Davis, T. P.; Fane, A. G. J. Mater. Chem. 2003, 13 (9), 2090−2097. (31) Hernández-Guerrero, M.; Davis, T. P.; Barner-Kowollik, C.; Stenzel, M. H. Eur. Polym. J. 2005, 41 (10), 2264−2277. (32) Fleet, R.; McLeary, J. B.; Grumel, V.; Weber, W. G.; Matahwa, H.; Sanderson, R. D. Eur. Polym. J. 2008, 44 (9), 2899−2911. (33) Hiltunen, M.; Riihelä, S.; Maunu, S. L. J. Polym. Sci., Part B: Polym. Phys. 2009, 47 (19), 1869−1879. (34) Semsarilar, M.; Ladmiral, V.; Perrier, S. J. Polym. Sci., Part A: Polym. Chem. 2010, 48 (19), 4361−4365. (35) Ifuku, S.; Kadla, J. F. Biomacromolecules 2008, 9 (11), 3308− 3313. (36) Hiltunen, M. S.; Raula, J.; Maunu, S. L. Polym. Int. 2011, 60 (9), 1370−1379. (37) Hiltunen, M.; Siirilä, J.; Aseyev, V.; Maunu, S. L. Eur. Polym. J. 2012, 48 (1), 136−145. (38) Hiltunen, M.; Siirilä, J.; Maunu, S. L. J. Polym. Sci., Part A: Polym. Chem. 2012, 50 (15), 3067−3076.

(39) Sui, X.; Yuan, J.; Zhou, M.; Zhang, J.; Yang, H.; Yuan, W.; Wei, Y.; Pan, C. Biomacromolecules 2008, 9 (10), 2615−2620. (40) Meng, T.; Gao, X.; Zhang, J.; Yuan, J.; Zhang, Y.; He, J. Polymer 2009, 50 (2), 447−454. (41) Xin, T.-T.; Yuan, T.; Xiao, S.; He, Y. BioResources 2011, 6 (3), 2941−2953. (42) Liu, Y.; Yao, K.; Chen, X.; Wang, J.; Wang, Z.; Ploehn, H. J.; Wang, C.; Chu, F.; Tang, C. Polym. Chem. 2014, 5, 3170−3181. (43) Jiang, F.; Wang, Z.; Qiao, Y.; Wang, Z.; Tang, C. Macromolecules 2013, 46 (12), 4772−4780. (44) Lin, C.; Zhan, H.; Liu, M.; Habibi, Y.; Fu, S.; Lucia, L. A. J. Appl. Polym. Sci. 2013, 127 (6), 4840−4849. (45) Hocine, S.; Li, M.-H. Soft Matter 2013, 9 (25), 5839−5861. (46) Liu, R.; Fraylich, M.; Saunders, B. Colloid Polym. Sci. 2009, 287 (6), 627−643. (47) Huijbrechts, A. A. M. L.; Huang, J.; Schols, H. A.; van Lagen, B.; Visser, G. M.; Boeriu, C. G.; Sudhölter, E. J. R. J. Polym. Sci., Part A: Polym. Chem. 2007, 45 (13), 2734−2744. (48) Strlič, M.; Kolenc, J.; Kolar, J.; Pihlar, B. J. Chromatogr. A 2002, 964 (1−2), 47−54. (49) Sjöholm, E.; Gustafsson, K.; Eriksson, B.; Brown, W.; Colmsjö, A. Carbohydr. Polym. 2000, 41 (2), 153−161. (50) Parry, K. L.; Shard, A. G.; Short, R. D.; White, R. G.; Whittle, J. D.; Wright, A. Surf. Interface Anal. 2006, 38 (11), 1497−1504. (51) Scofield, J. H. J. Electron Spectrosc. Relat. Phenom. 1976, 8 (2), 129−137. (52) Tanuma, S.; Powell, C. J.; Penn, D. R. Surf. Interface Anal. 1994, 21 (3), 165−176. (53) Tischer, T.; Goldmann, A. S.; Linkert, K.; Trouillet, V.; Börner, H. G.; Barner-Kowollik, C. Adv. Funct. Mater. 2012, 22 (18), 3853− 3864. (54) Kaupp, M.; Tischer, T.; Hirschbiel, A. F.; Vogt, A. P.; Geckle, U.; Trouillet, V.; Hofe, T.; Stenzel, M. H.; Barner-Kowollik, C. Macromolecules 2013, 46 (17), 6858−6872. (55) Gericke, M.; Schaller, J.; Liebert, T.; Fardim, P.; Meister, F.; Heinze, T. Carbohydr. Polym. 2012, 89 (2), 526−536. (56) Skey, J.; O’Reilly, R. K. Chem. Commun. 2008, 35, 4183−4185. (57) Copeland, C.; Conway, R.; Patroni, J.; Stick, R. Aust. J. Chem. 1981, 34 (3), 555−557. (58) Boyer, C.; Bulmus, V.; Priyanto, P.; Teoh, W. Y.; Amal, R.; Davis, T. P. J. Mater. Chem. 2009, 19 (1), 111−123. (59) Chaffey-Millar, H.; Stenzel, M. H.; Davis, T. P.; Coote, M. L.; Barner-Kowollik, C. Macromolecules 2006, 39 (19), 6406−6419. (60) Thomas, D. B.; Convertine, A. J.; Myrick, L. J.; Scales, C. W.; Smith, A. E.; Lowe, A. B.; Vasilieva, Y. A.; Ayres, N.; McCormick, C. L. Macromolecules 2004, 37 (24), 8941−8950. (61) Léonard, D.; Chevolot, Y.; Bucher, O.; Haenni, W.; Sigrist, H.; Mathieu, H. J. Surf. Interface Anal. 1998, 26 (11), 793−799. (62) Schmidt, B. V. K. J.; Hetzer, M.; Ritter, H.; Barner-Kowollik, C. Macromol. Rapid Commun. 2013, 34 (16), 1306−1311.

2572

dx.doi.org/10.1021/bm500416m | Biomacromolecules 2014, 15, 2563−2572